Gene Signature for the Prediction of NF-kappaB Activity

Abstract

Methods for predicting NF-kappaB (NF-kB) activity in a tumor, and more particularly to methods for predicting survival and therapeutic outcome, and selecting therapy in subjects with tumors, e.g., adenocarcinomas, e.g., lung adenocarcinomas and melanomas.

Description

CLAIM OF PRIORITY

[0001] This application claims the benefit of U.S. Provisional Patent Application No. 61/554,314, filed Nov. 1, 2011, the entire contents of which are incorporated by reference herein.

TECHNICAL FIELD

[0003] This invention relates to methods for predicting NF-kappaB (NF-kB) activity in a tumor, and more particularly to genes and gene signatures that predict survival and therapeutic outcome in subjects with tumors, e.g., adenocarcinoma, e.g., lung adenocarcinoma or melanoma.

BACKGROUND

[0004] The NF-kB family of transcription factors plays a crucial role in many cellular responses. They exist as homodimers or heterodimers of 5 distinct proteins: p50, p52, p65/RelA, RelB and cRel (Li and Verma, 2002. Nat Rev Immunol 2:725-734; Hayden and Ghosh, 2004. Genes Dev 18:2195-2224). NF-kB activation typically occurs by nuclear translocation following inducible phosphorylation of inhibitory IkB proteins by the IKKa/b (IkB kinase) complex (Hayden and Ghosh, 2004. Genes Dev 18:2195-2224; Hoffmann, and Baltimore, 2006. Immunol Rev 210:171-186; Karin and Ben-Neriah, 2000. Annu Rev Immunol 18:621-663; Vallabhapurapu and Karin, 2009. Annu Rev Immunol 27:693-733). Activation of the major conventional or canonical subunits p50, p65/RelA and cRel by inflammatory cytokines such as TNFa and IL-1a/b requires IKKb while non-conventional or non-canonical p52 and RelB subunits require IKKa (Bonizzi and Karin, 2004. Trends Immunol 25:280-288; Hacker and Karin, 2006. Sci STKE 2006:re13). A multitude of functions have been attributed to canonical NF-kB subunits, which include roles in inflammation and immunity, as well as in cell proliferation and survival. A tumor-promoting function for NF-kB in lymphomas has been known for some time (Staudt, 2010. Cold Spring Harb Perspect Biol 2:a000109; Packham, 2008. Br J Haematol 143:3-15). NF-kB was implicated in the Ras pathway (Mayo et al., 1997. Science 278:1812-1815) but the role of NF-kB in solid malignancies, although suspected, was not clear. Recent studies in mice and human cell lines have defined a key role for NF-kB in K-Ras-induced lung cancer ((Barbie et al., 2009. Nature 462:108-112; Meylan et al., 2009. Nature 462:104-107; Basseres et al., 2010. Cancer Res 70:3537-3546). This is likely through NF-kB activation via IKKb and/or TBK1 kinase by oncogenic K-Ras (Barbie et al., 2009. Nature 462:108-112; Meylan et al., 2009. Nature 462:104-107; Basseres et al., 2010. Cancer Res 70:3537-3546). It is not known whether NF-kB plays a general or genetic mutation-specific role in lung cancer development. An additional key role of IKKb/NF-kB in inflammation-promoting non-tumor myeloid cell types has also been shown to be critical in solid malignancies (Maeda et al., 2005. Cell 121:977-990; Karin and Greten, 2005. Nat Rev Immunol 5:749-759; Greten et al., 2004. Cell 118:285-296; Takahashi et al., 2010. Cancer Cell 17:89-97).

[0005] Inhibition of NF-kB activation represents a promising avenue for therapeutic targeting of lung cancer. However, relatively little is known about the role and activation state of NF-kB in human lung cancer. Most importantly, the inter-relation between NF-kB activation state and disease progression and survival is not known. It is also not known whether NF-kB is only activated in response to specific genetic mutations, e.g., K-Ras mutations. In this respect, one of the main stumbling blocks is the lack of an appropriate functional readout of NF-kB activation in human lung cancer cells. Previous NF-kB signatures have been defined, but not in lung cancer cells (Hinata et al., 2003. Oncogene 22:1955-1964; Boehm et al., 2007. Cell 129:1065-1079; Hernandez et al., 2010. Cancer Res 70:4005-4014).

SUMMARY

[0006] The present invention is based, at least in part, on the discovery of a set of genes that can be used to predict NF-kB activity, survival, and outcome in cancers, e.g., carcinoma, e.g., adenocarcinoma, e.g., human lung adenocarcinoma or melanoma.

[0007] Thus, in a first aspect, the invention provides methods for predicting NF-kB transcriptional activity in a tumor. The methods include determining gene expression levels for guanylate binding protein 1, interferon-inducible (GBP1); proteasome (prosome, macropain) subunit, beta type, 9 (large multifunctional peptidase 2) (PSMB9); interferon regulatory factor 1 (IRF1); transporter 1, ATP-binding cassette, sub-family B (MDR/TAP) (TAP 1); tumor necrosis factor, alpha-induced protein 3 (TNFAIP3); chemokine (C-C motif) ligand 5 (CCL5); proteasome (prosome, macropain) subunit, beta type, 8 (large multifunctional peptidase 7) (PSMB8); interleukin 32 (IL32); SH2B adaptor protein 3 (SH2B3); and nuclear factor of kappa light polypeptide gene enhancer in B-cells inhibitor, epsilon (NFKBIE) in a sample comprising cells from the tumor; and

comparing the gene expression levels to reference levels; optionally assigning a score to the tumor based on the comparison of the gene expression levels in the tumor to the reference levels; wherein the presence of gene expression levels above the reference levels, or a score above a threshold score, indicates that the tumor has high levels of NF-kB activity, and the presence of gene expression levels below the reference levels, or a score below the threshold score indicates that the tumor has low levels of NF-kB activity.

[0008] In some embodiments, the methods include identifying a subject as having a tumor having high levels of NF-kB activity, and selecting a treatment for the subject comprising administering an NF-kB inhibitor or immunotherapy.

[0009] In some embodiments, the methods include identifying a subject as having a tumor having low levels of NF-kB activity, and selecting a treatment for the subject comprising administering an NF-kB activator and immunotherapy.

[0010] In another aspect, the invention provides methods for selecting a treatment for a subject who has a tumor. The methods include determining gene expression levels for GBP1, PSMB9, IRF1, TAP1, TNFAIP3, CCL5, PSMB8, IL32, SH2B3 and NFKBIE in a sample comprising cells from the tumor; and

comparing the gene expression levels to reference levels; optionally assigning a score to the tumor based on the comparison of the gene expression levels in the tumor to the reference levels; detecting the presence of gene expression levels above the reference levels, or a score above a threshold score, and selecting a treatment for the subject comprising administering an NF-kB inhibitor or immunotherapy; or detecting the presence of gene expression levels below the reference levels, or a score below a threshold score, and selecting a treatment for the subject comprising administering an NF-kB activator and immunotherapy.

[0011] In another aspect, the invention provides methods for treating a subject who has a tumor. The methods include determining gene expression levels for GBP1, PSMB9, IRF1, TAP1, TNFAIP3, CCL5, PSMB8, IL32, SH2B3 and NFKBIE in a sample comprising cells from the tumor; and

comparing the gene expression levels to reference levels; optionally assigning a score to the tumor based on the comparison of the gene expression levels in the tumor to the reference levels; and detecting the presence in the subject of gene expression levels above the reference levels, or a score above a threshold score, and administering a treatment to the subject comprising an NF-kB inhibitor or immunotherapy, or detecting the presence in the subject of gene expression levels below the reference levels, or a score below a threshold score, and administering a treatment to the subject comprising an NF-kB activator and immunotherapy.

[0012] In another aspect, the invention provides methods for predicting outcome in a subject with a tumor. The methods include determining gene expression levels for GBP1, PSMB9, IRF1, TAP1, TNFAIP3, CCL5, PSMB8, IL32, SH2B3 and NFKBIE

in a sample comprising cells from the tumor; and comparing the gene expression levels to reference levels; optionally assigning a score to the tumor based on the comparison of the gene expression levels in the tumor to the reference levels; wherein the presence of gene expression levels below the reference levels, or a score below a threshold score, indicate that the subject is likely to have a poor outcome as compared to a subject who has levels above the reference levels or threshold score, and the presence of gene expression levels above the reference levels, or a score above a threshold score, indicate that the subject is likely to have a better outcome as compared to a subject who has levels below the reference levels or threshold score.

[0013] In another aspect, the invention provides methods for predicting outcome in a subject with a tumor. The methods include determining gene expression levels for some or all of CXCL1, CXCL2, CXCL3, IL6, and IL8, e.g., CXCL1, CXCL3, IL6, and IL8, in a sample comprising cells from the tumor; and

comparing the gene expression levels to reference levels; optionally assigning a score to the tumor based on the comparison of the gene expression levels in the tumor to the reference levels; wherein the presence of gene expression levels above the reference levels, or a score above a threshold score, indicate that the subject is likely to have a poor outcome as compared to a subject who has levels below the reference levels or threshold score, and the presence of gene expression levels below the reference levels, or a score above a threshold score, indicate that the subject is likely to have a better outcome as compared to a subject who has levels below the reference levels or threshold score.

[0014] In another aspect, the invention provides methods for predicting outcome in a subject with a tumor. The methods include determining gene expression levels of LTB in a sample comprising cells from the tumor; and

comparing the gene expression levels to reference levels; optionally assigning a score to the tumor based on the comparison of the gene expression levels in the tumor to the reference levels; wherein the presence of gene expression levels below the reference levels, or a score below a threshold score, indicate that the subject is likely to have a poor outcome as compared to a subject who has levels above the reference levels or threshold score, and the presence of gene expression levels above the reference levels, or a score above a threshold score, indicate that the subject is likely to have a better outcome as compared to a subject who has levels below the reference levels or threshold score.

[0015] In another aspect, the invention provides methods for monitoring treatment in a subject who has a tumor with low NF-kB activity. The methods include determining gene expression levels for GBP1, PSMB9, IRF1, TAP1, TNFAIP3, CCL5, PSMB8, IL32, SH2B3 and NFKBIE in a first sample comprising cells from the tumor; and

comparing the gene expression levels in the first sample to reference levels; optionally assigning a score to the tumor based on the comparison of the gene expression levels in the first sample to the reference levels; identifying a subject as having a tumor having low levels of NF-kB activity based on the presence of gene expression levels below the reference levels, or a score below the threshold score; administering to the subject a treatment, preferably a treatment comprising administering one or both of an NF-kB activator and immunotherapy; determining gene expression levels for GBP1, PSMB9, IRF1, TAP1, TNFAIP3, CCL5, PSMB8, IL32, SH2B3 and NFKBIE in a subsequent sample comprising cells from the tumor; and comparing the gene expression levels in the subsequent sample to levels in the first sample, wherein an increase in the levels from the first sample to the subsequent sample indicates that the treatment has been effective, and no change or a decrease in the levels from the first sample to the subsequent samples indicates that the treatment has not been effective.

[0016] In some embodiments, the tumor is a carcinoma, e.g., adenocarcinoma, e.g., lung adenocarcinoma or melanoma.

[0017] In some embodiments, determining gene expression levels comprises performing an assay to determine gene expression levels in the sample.

[0018] In some embodiments, assigning a score to the tumor based on the comparison of the gene expression levels in the tumor to the reference levels comprises using an algorithm to calculate a score.

[0019] In some embodiments, the NF-kB activator is an anticancer agent, preferably selected from the group consisting of taxanes, vinca alkaloids, and topoisomerase inhibitors.

[0020] In some embodiments, the NF-kB inhibitor is selected from the group consisting of sulfasalazine, Luteolin, rapamycin, temsirolimus and everolimus, caffeic acid phenethylester, SN50, parthenolide, triptolide, wedelolactone, lactacystin, substituted resorcinols, (E)-3-(4-methylphenylsulfonyl)-2-propenenitrile, Bay 11-7082, Bay 11-7821, or Bay 11-7085, Pranlukast, etoposide, bortezomib, MLN9708, PS-1145, tetrahydrocurcuminoids, such as Tetrahydrocurcuminoid CG, extracts of Paulownia tomentosa wood, and MG-132 (Z-Leu-Leu-Leu-H).

[0021] In some embodiments, the immunotherapy is selected from the group consisting of administration of dendritic cells or peptides with adjuvant; DNA-based vaccines; cytokines (e.g., IL-2); cyclophosphamide; anti-interleukin-2R immunotoxins; and antibodies, virus-based vaccines (e.g., adenovirus), formulations of Toll-like Receptor or RIG-I-like receptor ligands, Adoptive T cell therapy or other cell types.

[0022] In some embodiments, the antibodies are selected from the group consisting of anti-CD137, anti-PD1, anti-CD40, anti-PDL1, and anti-CTLA-4 antibodies.

[0023] In another aspect, the invention provides methods for predicting NF-kB transcriptional activity in a tumor. The methods include determining gene expression levels for some or all of the genes shown herein, e.g., in Tables A or 1 or 3, or the ten genes GBP1, PSMB9, IRF1, TAP1, TNFAIP3, CCL5, PSMB8, IL32, SH2B3 and NFKBIE, in a sample comprising cells from the tumor; comparing the gene expression levels to reference levels; and assigning a score to the tumor based on the comparison of the gene expression levels in the tumor to the reference levels. A score above a threshold score indicates that the tumor has high levels of NF-kB activity (and conversely, a score below the threshold indicates low NF-kB activity).

[0024] In some embodiments, the methods include identifying a subject as having a tumor having high levels of NF-kB activity, and selecting a treatment for the subject comprising administering an NF-kB inhibitor or immunotherapy.

[0025] In a further aspect, the invention provides methods of predicting outcome in a subject with cancer. The methods include determining gene expression levels for some or all of CXCL1, CXCL2, CXCL3, and IL8 in a sample comprising cells from the tumor; comparing the gene expression levels to reference levels; and optionally assigning a score to the tumor based on the comparison of the gene expression levels in the tumor to the reference levels. The presence of gene expression levels above the reference levels indicate that the subject is likely to have a poor outcome (and conversely, levels below the reference levels indicate a good outcome).

[0026] In some embodiments, the methods include identifying a subject as having a tumor having high gene expression levels, and selecting a treatment for the subject comprising administering an NF-kB inhibitor or immunotherapy.

[0027] In yet another aspect, the invention features methods for predicting outcome in a subject with cancer, The methods include determining gene expression levels of LTB in a sample comprising cells from the tumor; comparing the gene expression levels to reference levels; and optionally assigning a score to the tumor based on the comparison of the gene expression levels in the tumor to the reference levels. The presence of gene expression levels above the reference levels indicate that the subject is likely to have an improved outcome.

[0028] In some embodiments, the methods include identifying a subject as having a tumor having high LTB gene expression levels, and selecting a treatment for the subject comprising administering an NF-kB inhibitor or immunotherapy.

[0029] In another aspect, the invention features methods for predicting outcome in a subject with cancer. The methods include determining gene expression levels for some or all of CCL2, CCL5, LTB, CD83, and RELB in a sample comprising cells from the tumor; and comparing the gene expression levels to reference levels; optionally assigning a score to the tumor based on the comparison of the gene expression levels in the tumor to the reference levels; wherein the presence of gene expression levels above the reference levels indicate that the subject is likely to have an improved outcome.

[0030] In some embodiments, the methods include identifying a subject as having a tumor having high gene expression levels, and selecting a treatment for the subject comprising administering an NF-kB inhibitor or immunotherapy.

[0031] In some embodiments of the methods described herein, the tumor is a lung adenocarcinoma.

TABLE-US-00001 TABLE A GenBank ID GeneSymbol Gene description NM_000214.1 JAG1 jagged 1 (Alagille syndrome) NM_001166.3 BIRC2 baculoviral IAP repeat-containing 2 NM_001165 BIRC3 baculoviral IAP repeat-containing 3 NM_005504.4 BCAT1 branched chain aminotransferase 1, cytosolic NM_001710.4 BF B-factor, properdin NM_000064.1 C3 complement component 3 NM_001025079.1 CD47 CD47 antigen (Rh-related antigen, integrin- associated signal transducer) NM_004079.3 CTSS cathepsin S NM_015247 CYLD cylindromatosis (turban tumor syndrome) NM_001924.2 GADD45A growth arrest and DNA-damage-inducible, alpha NM_005238.2 ETS1 v-ets erythroblastosis virus E26 oncogene homolog 1 (avian) NM_002053.1 GBP1 guanylate binding protein 1, interferon- inducible, 67 kDa NM_001024070.1 GCH1 GTP cyclohydrolase 1 (dopa-responsive dystonia) NM_004004.3 GJB2 gap junction protein, beta 2, 26 kDa (connexin 26) NM_001511.1 CXCL1 chemokine (C-X-C motif) ligand 1 (melanoma growth stimulating activity, alpha) NM_002089.1 CXCL2 chemokine (C-X-C motif) ligand 2 NM_002090.2 CXCL3 chemokine (C-X-C motif) ligand 3 NM_005514.5 HLA-B major histocompatibility complex, class I, B AY732487.1 HLA-C major histocompatibility complex, class I, C NM_000201.1 ICAM1 intercellular adhesion molecule 1 (CD54), human rhinovirus receptor NM_000600.1 IL6 interleukin 6 (interferon, beta 2) NM_002184 IL6ST interleukin 6 signal transducer (gp130, oncostatin M receptor) NM_000584.2 IL8 interleukin 8 NM_001561.4 TNFRSF9 tumor necrosis factor receptor superfamily, member 9 NM_002192.2 INHBA inhibin, beta A (activin A, activin AB alpha polypeptide) NM_001570.3 IRAK2 interleukin-1 receptor-associated kinase 2 NM_002198.1 IRF1 interferon regulatory factor 1 NM_002203.2 ITGA2 integrin, alpha 2 (CD49B, alpha 2 subunit of VLA-2 receptor) NM_002205.2 ITGA5 integrin, alpha 5 (fibronectin receptor, alpha polypeptide) NM_000632.3 ITGAM integrin, alpha M (complement component receptor 3, alpha; also known as CD11b (p170), macrophage antigen alpha polypeptide) AU144005 ITGAV integrin, alpha V (vitronectin receptor, alpha polypeptide, antigen CD51) NM_005562.1 LAMC2 laminin, gamma 2 NM_009588.1 LTB lymphotoxin beta (TNF superfamily, member 3) NM_004994.2 MMP9 matrix metallopeptidase 9 (gelatinase B, 92 kDa gelatinase, 92 kDa type IV collagenase) NM_003998.2 NFKB1 nuclear factor of kappa light polypeptide gene enhancer in B-cells 1 (p105) NM_002502.2 NFKB2 nuclear factor of kappa light polypeptide gene enhancer in B-cells 2 (p49/p100) NM_004556.2 NFKBIE nuclear factor of kappa light polypeptide gene enhancer in B-cells inhibitor, epsilon NM_002526 NT5E 5'-nucleotidase, ecto (CD73) NM_002581.3 PAPPA pregnancy-associated plasma protein A, pappalysin 1 NM_198833.1 SERPINB8 serpin peptidase inhibitor, clade B (ovalbumin), member 8 NM_001005376.1 PLAUR plasminogen activator, urokinase receptor NM_004159.4 PSMB8 proteasome (prosome, macropain) subunit, beta type, 8 (large multifunctional peptidase 7) NM_002800.4 PSMB9 proteasome (prosome, macropain) subunit, beta type, 9 (large multifunctional peptidase 2) NM_002818.2 PSME2 proteasome (prosome, macropain) activator subunit 2 (PA28 beta) NM_002852.2 PTX3 pentraxin-related gene, rapidly induced by IL- 1 beta NM_006509.2 RELB v-rel reticuloendotheliosis viral oncogene homolog B, nuclear factor of kappa light polypeptide gene enhancer in B-cells 3 (avian) NM_002941.2 ROBO1 roundabout, axon guidance receptor, homolog 1 (Drosophila) NM_002982.3 CCL2 chemokine (C-C motif) ligand 2 NM_002985.2 CCL5 chemokine (C-C motif) ligand 5 NM_004591.1 CCL20 chemokine (C-C motif) ligand 20 NM_000593.5 TAP1 transporter 1, ATP-binding cassette, sub- family B (MDR/TAP) NM_000544.3 TAP2 transporter 2, ATP-binding cassette, sub- family B (MDR/TAP) NM_003190.3 TAPBP TAP binding protein (tapasin) NM_003264.3 TLR2 toll-like receptor 2 NM_006290.2 TNFAIP3 tumor necrosis factor, alpha-induced protein 3 NM_005658.3 TRAF1 TNF receptor-associated factor 1 NM_003300.2 TRAF3 TNF receptor-associated factor 3 NM_003916 AP1S2 adaptor-related protein complex 1, sigma 2 subunit NM_021101.3 CLDN1 claudin 1 NM_001012635.1 IL32 interleukin 32 NM_001040280.1 CD83 CD83 antigen (activated B lymphocytes, immunoglobulin superfamily) NM_004235.3 KLF4 Kruppel-like factor 4 (gut) NM_206853.1 QKI quaking homolog, KH domain RNA binding (mouse) NM_004289 NFE2L3 nuclear factor (erythroid-derived 2)-like 3 NM_014840.2 NUAK1 NUAK family, SNF1-like kinase, 1 NM_005110.1 GFPT2 glutamine-fructose-6-phosphate transaminase 2 NM_005124 NUP153 nucleoporin 153 kDa NM_005475.1 SH2B3 SH2B adaptor protein 3 NM_005493 RANBP9 RAN binding protein 9 NM_005729.3 PPIF peptidylprolyl isomerase F (cyclophilin F) NM_001008211.1 OPTN optineurin NM_006058.2 TNIP1 TNFAIP3 interacting protein 1 NM_006470.3 TRIM16 tripartite motif-containing 16 NM_006662 SRCAP Snf2-related CBP activator protein NM_017585.2 SLC2A6 solute carrier family 2 (facilitated glucose transporter), member 6 BC067106.1 GPR176 G protein-coupled receptor 176 NM_016445.1 PLEK2 pleckstrin 2 NM_015714.2 G0S2 G0/G1 switch 2 NM_016584.2 IL23A interleukin 23, alpha subunit p19 NM_001040458.1 ARTS-1 type 1 tumor necrosis factor receptor shedding aminopeptidase regulator NM_178031.2 HSPA5BP1 heat shock 70 kDa protein 5 (glucose-regulated protein, 78 kDa) binding protein 1 NM_018370.1 DRAM damage-regulated autophagy modulator NM_018351.2 FGD6 FYVE, RhoGEF and PH domain containing 6 NM_020183.3 ARNTL2 aryl hydrocarbon receptor nuclear translocator- like 2 XM_934503.1 FAM91A2 family with sequence similarity 91, member A2 NM_022154.5 SLC39A8 solute carrier family 39 (zinc transporter), member 8 NM_022168.2 IFIH1 interferon induced with helicase C domain 1 NM_022350.1 ERAP2 endoplasmic reticulum aminopeptidase 2 NM_022750.2 PARP12 poly (ADP-ribose) polymerase family, member 12 NM_022763.2 FNDC3B Caution, check this probeset carefully. This probeset may detect an alternate exon, an alternate termination site, or an overlapping transcript of fibronectin type III domain containing 3B NM_024615.2 PARP8 poly (ADP-ribose) polymerase family, member 8 NM_030952.1 NUAK2 NUAK family, SNF1-like kinase, 2 NM_031449.3 ZMIZ2 zinc finger, MIZ-type containing 2 NM_032413.2 C15orf48 chromosome 15 open reading frame 48 NM_014903.3 NAV3 neuron navigator 3 NM_138397 LOC93082 hypothetical protein BC012317 NM_030968.2 C1QTNF1 C1q and tumor necrosis factor related protein 1 NM_173490 LOC134285 hypothetical protein LOC134285 NM_001031739.1 ASB9 ankyrin repeat and SOCS box-containing 9 NM_178496.2 C3orf59 chromosome 3 open reading frame 59 NM_144975 MGC19764 hypothetical protein MGC19764 NM_173545.1 C2orf13 chromosome 2 open reading frame 13 NM_147156.3 TMEM23 transmembrane protein 23 NM_207376.1 OCC-1 overexpressed in colon carcinoma-1 XM_498811.2 KIAA0493 KIAA0493 protein XM_930678.1 LOC642441 hypothetical LOC642441 XM_937100.1 LOC728285 similar to keratin associated protein 2-4 NM_002223.2 ITPR2 inositol 1,4,5-triphosphate receptor, type 2 NM_002192.2 INHBA Caution, this probeset may detect an extended transcript or alternative terminal exon for inhibin, beta A (activin A, activin AB alpha polypeptide) AK000776 ROR1 Caution, this probeset may detect an extended transcript or alternative terminal exon for receptor tyrosine kinase-like orphan receptor 1 AU144005 ITGAV Caution, this probeset may detect an alternative exon for integrin, alpha V (vitronectin receptor, alpha polypeptide, antigen CD51) NM_000382.2 ALDH3A2 aldehyde dehydrogenase 3 family, member A2 NM_020987.2 ANK3 ankyrin 3, node of Ranvier (ankyrin G) NM_004058.2 CAPS calcyphosine NM_001752.2 CAT catalase NM_001875.2 CPS1 carbamoyl-phosphate synthetase 1, mitochondrial NM_004390.2 CTSH cathepsin H NM_001352.2 DBP D site of albumin promoter (albumin D-box) binding protein NM_001005336.1 DNM1 dynamin 1 NM_004409.2 DMPK dystrophia myotonica-protein kinase NM_001958.2 EEF1A2 eukaryotic translation elongation factor 1 alpha 2 NM_001408.1 CELSR2 cadherin, EGF LAG seven-pass G-type receptor 2 (flamingo homolog, Drosophila) NM_177996.1 EPB41L1 erythrocyte membrane protein band 4.1-like 1 NM_005252.2 FOS v-fos FBJ murine osteosarcoma viral oncogene homolog NM_002072 GNAQ guanine nucleotide binding protein (G protein), q polypeptide NM_002081.1 GPC1 glypican 1 NM_005896.2 IDH1 isocitrate dehydrogenase 1 (NADP+), soluble NM_002168.2 IDH2 isocitrate dehydrogenase 2 (NADP+), mitochondrial NM_021070.2 LTBP3 latent transforming growth factor beta binding protein 3 NM_002374.3 MAP2 microtubule-associated protein 2 NM_001012333.1 MDK midkine (neurite growth-promoting factor 2) NM_000435.1 NOTCH3 Notch homolog 3 (Drosophila) NM_002585 PBX1 pre-B-cell leukemia transcription factor 1 NM_000920 PC pyruvate carboxylase NM_005391.1 PDK3 pyruvate dehydrogenase kinase, isozyme 3 NM_002705.3 PPL periplakin NM_002737.2 PRKCA protein kinase C, alpha NM_152880.2 PTK7 PTK7 protein tyrosine kinase 7 NM_001038.4 SCNN1A sodium channel, nonvoltage-gated 1 alpha NM_003355.2 UCP2 uncoupling protein 2 (mitochondrial, proton carrier) NM_003389 CORO2A coronin, actin binding protein, 2A NM_003568.1 ANXA9 annexin A9 NM_003786.2 ABCC3 ATP-binding cassette, sub-family C (CFTR/MRP), member 3 NM_004753.4 DHRS3 dehydrogenase/reductase (SDR family) member 3 NM_005117.2 FGF19 fibroblast growth factor 19 NM_005759 ABI-2 abl-interactor 2 NM_002510 GPNMB glycoprotein (transmembrane) nmb NM_006393 NEBL nebulette NM_007365 PADI2 peptidyl arginine deiminase, type II NM_014988.1 LIMCH1 LIM and calponin homology domains 1 NM_015132 SNX13 sorting nexin 13 NM_015271.2 TRIM2 tripartite motif-containing 2 NM_014467.1 SRPX2 sushi-repeat-containing protein, X-linked 2 NM_014067.2 MACROD1 MACRO domain containing 1 NM_013296.3 GPSM2 G-protein signalling modulator 2 (AGS3-like, C. elegans) NM_016619.1 PLAC8 placenta-specific 8 NM_016233.1 PADI3 peptidyl arginine deiminase, type III AK096661.1 DKFZP761M1511 hypothetical protein DKFZP761M1511 NM_019027.1 RBM47 RNA binding motif protein 47 AF172820.1 MXRA6 matrix-remodelling associated 6 NM_032379.3 SYTL2 synaptotagmin-like 2 NM_017734.2 PALMD palmdelphin NM_017905.3 TMCO3 transmembrane and coiled-coil domains 3 NM_020169.2 LXN latexin NM_020397.2 CAMK1D calcium/calmodulin-dependent protein kinase ID AB007969.1 CLMN calmin (calponin-like, transmembrane) NM_021180.2 GRHL3 grainyhead-like 3 (Drosophila) NM_022783.1 DEPDC6 DEP domain containing 6 NM_024539.3 RNF128 ring finger protein 128 NM_024896.2 KIAA1815 KIAA1815 NM_025094 FLJ22184 hypothetical protein FLJ22184 NM_024491.2 CEP70 centrosomal protein 70 kDa NM_024491 BITE p10-binding protein NM_030821 PLA2G12 phospholipase A2, group XII NM_032042.3 C5orf21 chromosome 5 open reading frame 21 NM_032229.2 SLITRK6 SLIT and NTRK-like family, member 6 NM_032872.1 SYTL1 synaptotagmin-like 1 NM_177963.2 SYT12 synaptotagmin XII NM_138393.1 C19orf32 chromosome 19 open reading frame 32 NM_080597.2 OSBPL1A oxysterol binding protein-like 1A NM_199165.1 ADSSL1 adenylosuccinate synthase like 1

NM_138962 MSI2 musashi homolog 2 (Drosophila) NM_138801 LOC130589 aldose 1-epimerase NM_138801.1 GALM galactose mutarotase (aldose 1-epimerase) NM_144658.2 DOCK11 dedicator of cytokinesis 11 NM_152527.3 SLC16A14 solute carrier family 16 (monocarboxylic acid transporters), member 14 NM_170743.2 IL28RA interleukin 28 receptor, alpha (interferon, lambda receptor) NM_001012642.1 GRAMD2 GRAM domain containing 2 NM_174921.1 C4orf34 chromosome 4 open reading frame 34 NM_174921 LOC201895 hypothetical protein LOC201895 NM_207362.1 C2orf55 chromosome 2 open reading frame 55 NM_174921.1 C4orf34 chromosome 4 open reading frame 34 AA732944 RALGPS2 Ral GEF with PH domain and SH3 binding motif 2 AU145501 NAALADL2 N-acetylated alpha-linked acidic dipeptidase- like 2 NM_006252.3 PRKAA2 protein kinase, AMP-activated, alpha 2 catalytic subunit AW294903 WNT7B wingless-type MMTV integration site family, member 7B

[0032] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Methods and materials are described herein for use in the present invention; other, suitable methods and materials known in the art can also be used. The materials, methods, and examples are illustrative only and not intended to be limiting. All publications, patent applications, patents, sequences, database entries, and other references mentioned herein are incorporated by reference in their entirety. In case of conflict, the present specification, including definitions, will control.

[0033] Other features and advantages of the invention will be apparent from the following detailed description and figures, and from the claims.

DESCRIPTION OF DRAWINGS

[0034] FIGS. 1A-H. Impact of IKKb-induced NF-kB on tumor rejection. (a) EMSA showing NF-kB nuclear levels in LLC-OVA transduced with control MiG, IKK and MiG treated with TNFa for 1 h and 2 h as indicated. (b) RT-PCR showing KC/CXCL1 expression in LLC-OVA transduced with control MiG, IKK and MiG treated with TNFa for 1 h and 2 h as indicated. Samples run in triplicate and reported as mean+/-SEM. (c) Tumor growth in C57B1/6 mice inoculated s.c. with non-immunogenic LLC-MiG and LLC-IKK over indicated time periods. Each line represents a single mouse. (d) Tumor growth in C57B1/6 mice inoculated s.c. with immunogenic LLC-OVA-MiG and LLC-OVA-IKK over indicated time periods. Each line represents a single mouse. (e) Impact of immunogenic-LLC tumors on peripheral T cells. Tetramer analysis of OVA-specific CD8 T cells in peripheral blood on day 10 from naive mice or mice receiving LLC-OVA-MiG or LLC-OVA-IKK cells s.c. Each point represents a single mouse. Student's T test was performed to compare tetramer positive CD8 T cells between mice receiving LLC-OVA-MiG and LLC-OVA-IKK tumors. (f) C57B1/6 mice received s.c. LLC-OVA-MiG or LLC-OVA-IKK and tumor growth was monitored. Relative fold increase in tumor volume in mice at D21 post-inoculation compared to D4 post-inoculation. Combined results from 3 independent experiments are shown (n=11 for both groups). Each point represents tumor growth from a single mouse. (g) Tumor growth in RAG2-/- mice inoculated s.c. with LLC-OVA-MiG or LLC-OVA-IKK. Each line represents a single mouse. (h) Tumor growth in L129/sv mice inoculated s.c. with immunogenic LKR-OVA-MiG and LKR-OVA-IKK over indicated time periods. Each line represents a single mouse.

[0035] FIG. 1I. Impact of immunogenic-LLC tumors (LLC-OVA) on peripheral OVA-specific CD8 T cells. Tetramer analysis showing percent OVA-specific CD8 T cells in peripheral blood (out of total CD8 T cells) on day 10 from mice receiving LLC-OVA s.c or naive mice as indicated. Each point represents a single mouse. Student's T test was performed to compare tetramer positive CD8 T cells between naive mice and those receiving LLC-OVA tumors.

[0036] FIG. 1J. Activation of NF-kB in non-immunogenic LLC cells. EMSA showing nuclear levels of NF-kB (top) and AP1 (bottom) in LLC-MiG, LLC-IKK, or LLC-MiG cells treated with TNF.alpha.. Results are representative of at least two independent experiments.

[0037] FIG. 1K. NF-kB and AP1 activity as determined by EMSA in parental LKR-13, LKR-13 transduced with the control MiG retrovirus (MiG), IkB.alpha.SR retrovirus (IkB.alpha.) and CA-IKK.beta. retrovirus (IKK.beta.). Mobility of different complexes is indicated with arrows.

[0038] FIGS. 2A-B. NF-kB enhances tumor rejection in vaccine-activated T cell models. (a) Growth of LLC in a metastatic model of lung cancer. H&E staining of lungs from mice 24 days after receiving i.v. non-immunogenic LLC-MiG or LLC-IKK, or immunogenic LLC-OVA-MiG or LLC-OVA-IKK. All scale bars represent 4 mm. (b) C57B1/6-BALB/c F1 (CB6) mice received s.c. TUBO-MiG or TUBO-IKK. After 5 days, half of the mice in each group received HER2 TriVax and tumor growth was monitored. Tumor growth of all mice was calculated at D21 relative to D5 and relative growth in vaccinated mice was compared to their unvaccinated counterparts. p-value calculated using Student's T test with Welch's correction. Graph shows combined results of 2 independent experiments, each point represents tumor growth from a single mouse.

[0039] FIG. 2C. Presence of HER2-specific CD8 T cells after TriVax administration. Tetramer analysis of peripheral blood on day 10 from naive mice or mice receiving TUBO-MiG or TUBO-IKK with or without TriVax vaccination. Vaccination given on day 5. Each point represents data from a single mouse. Results are representative of two independent experiments. Student's T test was performed to compare tetramer positive CD8 T cells between TriVax treated mice receiving TUBO-MiG and those receiving TUBO-IKK tumors.

[0040] FIGS. 3A-D. Impact of T cell chemokines on tumor rejection. (a) IFNg production by OT-1 CD8 T cells. ELISpot of OT-I T cells cultured with LLC-parental, LLC-OVA, LLC-OVA-MiG, or LLC-OVA-IKK tumor cells. 1x105 LLC targets/well. 1x105 T cells/well. Samples were run in triplicate with results reported as mean+/-SEM. (b) Affymetrix probe set signal intensity of indicated chemokines in LLC-OVA-IKK compared to LLC-OVA-MiG. Genes identified in two separate microarray experiments are shown, and reported as mean+/-SEM. (c) RT-PCR showing CCL2 and CCL5 expression in LLC parental, LLC transduced with OVA and LLC transduced with OVA+MiG or OVA+IKK. Samples were run in triplicate and reported as average+/-SEM. (d) LLC-OVA-IKK-Lenti control and LLC-OVA-IKK-CCL2 KD cells were injected s.c. in C57B1/6 mice and tumor growth was monitored. Each line represents tumor growth in a single mouse. All results are representative of at least two independent experiments.

[0041] FIG. 3E. Impact of IKK.beta. expression on chemokine expression in LLC-OVA determined by RNA microarray analysis on an Affymetrix platform. Affymetrix probe set intensity fold increase in indicated chemokine expression in LLC-OVA-IKK compared to LLC-OVA-MiG. Genes identified in two separate microarray experiments are shown, and reported as mean+/-SEM.

[0042] FIG. 3F. LLC-OVA-MiG and LLC-OVA-IKK supernatants were collected 24 h after plating and the amount of secreted CCL2 was determined by ELISA.

[0043] FIGS. 4A-E. Generation and validation of a lung cancer NF-kB signature. (a) EMSA analysis of sorted H23, PC9 and HCC827 cells infected with GFP, IkB or IKKb retroviruses as indicated. Parental (Par) were not infected. The major NF-kB complexes are indicated with arrows. (b) Determination of mRNA expression of NF-kB target genes BIRC3 and TNFAIP3 in low NF-kB signature (H322, H1395, H522, H1437) and high NF-kB signature (H226, H157, H1299, H650) cells. Relative expression is shown after normalizing to 18s rRNA levels. Samples were run in triplicate and reported as average+/-SEM. Students t-test was used to determine statistical significance in difference in mean expression of BIRC3 and TNFAIP3 in low versus high cell lines. (c) H226, H157 and H1299 were transduced with MiG or IkBaSR retroviruses following which BIRC3 and TNFAIP3 mRNA expression was determined as in "b". (d) Expression of CCL2, CCL5, CXCL1-3 and IL8 was determined in H157 cells transduced with MiG and IkBaSR. (e) NF-kB activity determined by EMSA in NF-kB signature low (1-4) and high (5-8) cell lines. Major NF-kB complexes are indicated with arrows.

[0044] FIGS. 5A-D. Association of T cell chemokines, but not neutrophil chemokines, with T cell presence in human lung adenocarcinoma. (a) Correlation plot based on Spearman correlation r-value of CMCLA gene expression data in human adenocarcinomas (n=442) for T cell chemokines (CCL2, CCL5, CXCL10), neutrophil chemokines (CXCL1-3, IL8), and T cell receptor (TRAC, TRBC1) genes. Gene names and Affymetrix probe set ID numbers are shown. (b) Correlation r-values of LTb expression with T cell chemokines, neutrophil chemokines, and T cell presence in CMCLA dataset (n=442). Gene name and Affymetrix probe set ID numbers for genes with 2 probe sets are shown. (c) mRNA expression of indicated genes normalized to 18s rRNA in HCC827 lung cancer cells determined by RT-PCR. Fold difference in expression of genes after LT treatment compared to untreated cells is shown. Samples were run in triplicate and reported as mean+/-SEM. (d) mRNA expression of indicated genes normalized to 18s rRNA in HCC827 lung cancer cells determined by RT-PCR. Fold difference in expression of genes after TNFa treatment compared to untreated cells is shown. Samples were run in triplicate and reported as mean+/-SEM.

[0045] FIG. 5E. Correlation of expression between genes in human lung adenocarcinoma. Correlation r-values of expression of different NF-kB activating cytokine genes (y-axis) with expression of different genes (x-axis). Gene name and Affymetrix probe set numbers for genes with 2 probe sets are shown.

[0046] FIGS. 6A-D. Association of NF-kB signature inflammatory genes with patient survival. (a-d) Association of mRNA expression of indicated inflammatory genes (and Affymetrix probe set ID) with OS in CMCLA dataset (n=442). 5-year OS of patients exhibiting high versus low expression of indicated genes using a median cutoff. Kaplan-Meier method was used to generate survival curves and the log-rank test was used to test survival difference between the low and high expression groups by median cutoff for each gene. The p-value shown was adjusted by false discovery rate (FDR) for multiple testing.

[0047] FIGS. 7A-E. Association of NF-kB signature immune response genes and T cell presence with patient survival. (a-d) Association of mRNA expression of indicated immune response genes (and Affymetrix probe set ID) with OS in CMCLA (n=442). 5-year OS of patients exhibiting high versus low expression of indicated genes using a median cutoff. Kaplan-Meier method was used to generate survival curves and the log-rank test was used to test survival difference between the low and high expression groups by median cutoff for each gene. The p value shown was adjusted by false discovery rate (FDR) for multiple testing. (e) Association of T cell presence detected by expression of T cell receptor genes (TRAC, TRBC1) on patient survival was determined as with above genes.

[0048] FIG. 8A. Correlation of expression between each of the ten NF-.kappa.B driver genes (y-axis) and PC1 (top panels) and PC2 (bottom panels) of the NF-.kappa.B signature 159 probesets (x-axis) in the CMCLA dataset. Correlation r-values, gene names and Affymetrix probe set ID are shown.

[0049] FIG. 8B. The first 3 Principal Components (PC1-3) of the T cell receptor genes (TRAC and TRBC1) (Top panel) or the ten-gene NF-kB signature (bottom panel) are shown in the CMCLA dataset.

[0050] FIGS. 8C-G. Association of T cell presence with NF-kB activity in human lung cancer. (c) Spearman correlation plot with r-value of NF-kB signature (159 probe sets) PC1 with T cell PC1 are shown for CMCLA data (n=442). (d) Correlation plot of NF-kB signature (159 probe sets) PC1 with MR signature PC1. (e) Correlation plot of NF-kB signature PC1 (159 probe sets) with 10-gene NF-kB signature PC1. (f) Correlation plot of T cell PC1 with 10-gene NF-kB signature PC1. (g) Correlation plot of MR signature PC1 with 10-gene NF-kB signature PC1.

[0051] FIG. 8H. Correlation of expression between the 10-gene NF-kB signature with T cell presence and the MR signature in the GSE14814 dataset (n=133). Correlation r-values are shown.

[0052] FIG. 9. Correlation plot of NF-kB signature (159 probe sets) PC1 (top panel) or 10-gene NF-kB signature PC1 (bottom panel) with T cell presence in the melanoma dataset GSE19234. Correlation r-values are shown.

DETAILED DESCRIPTION

[0053] Previously, it was not known whether tumor NF-kB regulates T cell-mediated anti-tumor responses and immune surveillance. Nonetheless, consistent with known pro-tumor functions, it is possible that NF-kB also impairs anti-tumor T cell responses through cancer cell-intrinsic and/or microenvironment effects. However, as shown herein, in immunogenic tumors in mice NF-kB induces T cell-mediated tumor rejection. Enhanced T cell recruitment was found to be a key NF-kB dependent mechanism for tumor rejection. To investigate potential pro-tumor and anti-tumor NF-kB functions in human cancer, a novel human lung cancer NF-kB gene expression signature was developed. Although there was evidence of both inflammatory and immune-response functions, overall NF-kB activity was strongly associated with T cell presence in human lung cancer; as T cell presence in tumors can be associated with immune surveillance and improved patient survival, NF-kB activity as determined by the gene signature described herein is prognostic. These findings in both murine and human lung cancer indicate that a crucial and previously unappreciated function of tumor NF-kB is to promote T cell-mediated immune surveillance responses; thus, the NF-kB gene signature can be used to select subjects for treatment with immunotherapy.

[0054] As described herein, identification of genes that are regulated by NF-kB specifically in lung cancer cells provide an indicator of NF-kB activation state. Such an NF-kB signature can then be used to predict disease outcome and survival, and validate this pathway as a potentially crucial therapeutic target in human lung cancer. To this end, an NF-kB signature was established in lung cancer cell lines and the correlation between NF-kB activation state and disease outcome and patient survival was determined using the Consortium for the Molecular Classification of Lung Adenocarcinoma (CMCLA) survival prediction study (Shedden et al. 2008. Nat Med 14:822-827).

[0055] Methods of Treating, Assigning a Prognosis or Predicting Survival

[0056] The methods can be used to monitor a treatment (e.g., an immunotherapy or administration of an NF-kB inhibitor), or to select a treatment, e.g., to select a treatment regime including an immunotherapy or administration of an NF-kB inhibitor for a subject. In addition, the methods described herein can be used for, e.g., to assist in, assigning a prognosis or predicting survival in a subject who has a tumor, e.g., a solid tumor.

[0057] As used herein, the term "cancer" refers to cells having the capacity for autonomous growth, i.e., an abnormal state or condition characterized by rapidly proliferating cell growth. Hyperproliferative and neoplastic disease states may be categorized as pathologic, i.e., characterizing or constituting a disease state, or may be categorized as non-pathologic, i.e., a deviation from normal but not associated with a disease state. In general, a cancer will be associated with the presence of one or more tumors, i.e., abnormal cell masses. The term "tumor" is meant to include all types of cancerous growths or oncogenic processes, metastatic tissues or malignantly transformed cells, tissues, or organs, irrespective of histopathologic type or stage of invasiveness. "Pathologic hyperproliferative" cells occur in disease states characterized by malignant tumor growth. In general, the methods described herein can be practiced on subjects with solid tumors.

[0058] Tumors include malignancies of the various organ systems, such as affecting lung, breast, thyroid, lymphoid, gastrointestinal, and genito-urinary tract, as well as adenocarcinomas which include malignancies such as most colon cancers, renal-cell carcinoma, prostate cancer and/or testicular tumors, non-small cell carcinoma of the lung, cancer of the small intestine and cancer of the esophagus. The term "carcinoma" is art recognized and refers to malignancies of epithelial or endocrine tissues including respiratory system carcinomas, gastrointestinal system carcinomas, genitourinary system carcinomas, testicular carcinomas, breast carcinomas, prostatic carcinomas, endocrine system carcinomas, and melanomas. In some embodiments, the disease is renal carcinoma or melanoma. Exemplary carcinomas include those forming from tissue of the cervix, lung, prostate, breast, head and neck, colon and ovary. The term also includes carcinosarcomas, e.g., which include malignant tumors composed of carcinomatous and sarcomatous tissues. An "adenocarcinoma" refers to a carcinoma derived from glandular tissue or in which the tumor cells form recognizable glandular structures. The term "sarcoma" is art recognized and refers to malignant tumors of mesenchymal derivation.

[0059] In some embodiments, cancers evaluated by the methods described herein include those that are particularly immunogenic, e.g., neuroblastoma, melanoma, and renal cell cancer.

[0060] In some embodiments, cancers or tumors evaluated by the methods described herein include carcinomas (i.e., epithelial cancers), such as a lung cancer (e.g., non-small-cell lung cancer (NSCLC)), breast cancer, colorectal cancer, head and neck cancer, or ovarian cancer. Epithelial malignancies are cancers that affect epithelial tissues, and include adenocarcinomas and squamous cell carcinoma. In some embodiments, the methods can be used to evaluate or treat adenocarcinomas.

[0061] The present methods can include the use of some or all of the genes shown in Table A or Table 1, e.g., 10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 110, 120, 130, 138, 140, 150, 160, 170, 180, 190, or more of the genes in Table 1, e.g., the genes shown in Table 2 or 3, or more preferably the genes in the ten-gene signature (i.e., GBP1, PSMB9, IRF1, TAP1, TNFAIP3, CCL5, PSMB8, IL32, SH2B3, and NFKBIE), to predict NF-kB transcriptional activity in different tumor types, predict outcome in subjects with tumors, and select treatments for subjects with tumors. This can be done using methods known in the art, e.g., using a classifier based on principal component analysis (PCA), e.g., as described below; by a classifier based on weighted majority voting; or classifiers based on multi-dimensional scaling of the full 240 genes or a sub-set of the genes described in this application or in U.S. Ser. No. 61/554,314 (Shedden et al., 2008. Nat Med 14:822-827; Chen et al., 2010. Breast Cancer Res Treat 119:335-346). Those tumors found to have high levels of NF-kB activity can then be selected for treatment, e.g., with an NF-kB inhibitor, and NF-kB activator, and/or with immunotherapy. Depending on levels of NF-kB activity (high or low), subjects can be selected can be selected for treatment, e.g., with an NF-kB inhibitor or with immunotherapy (with high NF-kB activity) or compounds capable of enhancing NF-kB to enhance effect of immunotherapy (with low NF-kB activity).

[0062] Further, five inflammatory genes (CXCL1, CXCL3, IL6, and IL8) have been identified that predict poor outcome (see FIGS. 6A-D). These genes are thought to mediate inflammatory as well as tumor invasion and metastasis responses. The methods can include identifying subjects having tumors with high levels of expression of those genes, and treating them with NF-kB pathway inhibitors.

[0063] Additional NF-kB signature genes have been identified (CCL2, ICAM-1, LTB, and CD83) predict improved survival (see FIGS. 7A-D). High expression of these genes is expected to be associated with improved response to cancer immunotherapy treatment, thus, the methods can include identifying subjects having tumors with high levels of expression of those genes, and treating them with immunotherapy.

[0064] Finally, as demonstrated herein, LTB individually predicts improved outcome (FIG. 7B) and the presence of T cells in tumors. LTB induces expression of T cell chemokines such as CCL2 and CCL5 in lung cancer cells. High expression of LTB is expected to be associated with improved response to cancer immunotherapy treatment. Thus, the methods can include identifying subjects having tumors with high levels of expression of LTB, and treating them with immunotherapy.

[0065] The methods described herein generally include obtaining a sample from a subject, and evaluating the presence and/or level of an NF-kB signature gene in the sample, and comparing the presence and/or level with one or more references, e.g., a control reference that represents a normal level of the NF-kB signature gene, e.g., a level in an unaffected subject, a level in a normal, non-tumor tissue of the subject (e.g., normal lung tissues) and/or a disease reference that represents a level associated with outcome or survival. The methods can include detecting the level of a gene as described herein, or a protein encoded by a gene described herein. The presence and/or level of a gene or protein can be evaluated using methods known in the art, e.g., using quantitative PCR methods or molecular barcoding technologies, e.g., NanoString.TM. (see, e.g., U.S. Pat. No. 7,919,237; U.S. Pat. No. 7,473,767; and Geiss et al., Nature Biotechnology 26: 317-25 (2008)). In some embodiments, high throughput methods, e.g., protein or gene chips as are known in the art (see, e.g., Ch. 12, Genomics, in Griffiths et al., Eds. Modern genetic Analysis, 1999, W. H. Freeman and Company; Ekins and Chu, Trends in Biotechnology, 1999, 17:217-218; MacBeath and Schreiber, Science 2000, 289(5485):1760-1763; Simpson, Proteins and Proteomics: A Laboratory Manual, Cold Spring Harbor Laboratory Press; 2002; Hardiman, Microarrays Methods and Applications: Nuts & Bolts, DNA Press, 2003), can be used to detect the presence and/or level of an NF-kB signature gene or gene product.

[0066] In some embodiments, the methods include assaying levels of one or more control genes or proteins, and comparing the level of expression of the immune-related genes or proteins to the level of the control genes or proteins, to normalize the levels of the immune-related genes or proteins. Suitable endogenous control genes includes a gene whose expression level should not differ between samples, such as a housekeeping or maintenance gene, e.g., 18S ribosomal RNA; beta Actin; Glyceraldehyde-3-phosphate dehydrogenase; Phosphoglycerate kinase 1; Peptidylprolyl isomerase A (cyclophilin A); Ribosomal protein L13a; large Ribosomal protein P0; Beta-2-microglobulin; Tyrosine 3-monooxygenase/tryptophan 5-monooxygenase activation protein, zeta polypeptide; Succinate dehydrogenase; Transferrin receptor (p90, CD71); Aminolevulinate, delta-, synthase 1; Glucuronidase, beta; Hydroxymethyl-bilane synthase; Hypoxanthine phosphoribosyltransferase 1; TATA box binding protein; and/or Tubulin, beta polypeptide.

[0067] Generally speaking, the methods described herein can be performed on cells from a tumor, e.g., a benign or malignant tumor. The cells can be obtained by known methods, e.g., during a biopsy (such as a core needle biopsy), or during a surgical procedure to remove all or part of the tumor. The cells can be used fresh, frozen, fixed, and/or preserved, so long as the mRNA or protein that is to be assayed is maintained in a sufficiently intact state to allow accurate analysis.

[0068] In some embodiments of the methods described herein, the levels of the immune-related genes in the tumor sample can be compared individually to levels in a reference. The reference levels can represent levels in a subject who has a good outcome, a good prognosis, or a long predicted survival time (e.g., 2 years or more). Alternatively, reference levels can represent levels in a subject who has a poor prognosis, or a shorter predicted survival time (e.g., less than 2 years). In some embodiments, the reference levels represent a threshold, and a level in the tumor that is above the threshold reference level indicates that the subject has a good outcome, a good prognosis, or a long predicted survival time (e.g., 2 years or more), and levels below the threshold reference level indicates that the subject has a poor outcome, a poor prognosis, or a shorter predicted survival time (e.g., less than 2 years).

[0069] In some embodiments, the reference levels can represent levels in a subject who is predicted to respond to immunotherapy. Alternatively, reference levels can represent levels in a subject who is predicted to have no or a poor response to immunotherapy. In some embodiments, the reference levels represent a threshold, and a level in the tumor that is above the threshold reference level indicates that the subject is predicted to respond to immunotherapy, and levels below the threshold reference level indicates that the subject is predicted to have no or poor response to immunotherapy. In subjects who are predicted respond to immunotherapy, the methods can further include administering an immunotherapy for those subjects, or selecting or recommending a treatment including an immunotherapy for those subjects.

[0070] In some embodiments of the methods described herein, values representing the levels of the NF-kB genes (or subsets thereof, e.g., as described herein) can be summed to produce a "NF-kB gene score" that can be compared to a reference NF-kB gene score, wherein an NF-kB gene score that is above the reference NF-kB gene score indicates that the subject has a long predicted survival time (e.g., 2 years or more) or is predicted to have a positive response to immunotherapy, and an NF-kB gene score below the reference score indicates that the subject has a shorter predicted survival time (e.g., less than 2 years), or is predicted to have no or a poor response to immunotherapy. In subjects with an NF-kB gene score below the reference score, an NF-kB activator can be administered in conjunction with immunotherapy.

[0071] For example, in some embodiments, the expression levels of each of the evaluated genes can be assigned a value (e.g., a value that represents the expression level of the gene, e.g., normalized to an endogenous control gene as described herein). That value (optionally weighted to increase or decrease its effect on the final score) can be summed to produce an immune-related gene score. One of skill in the art could optimize such a method to determine an optimal algorithm for determining an NF-kB gene score.

[0072] One of skill in the art will appreciate that references can be determined using known epidemiological and statistical methods, e.g., by determining an NF-kB gene score, or NF-kB gene-encoded protein or mRNA levels, in tumors from an appropriate cohort of subjects, e.g., subjects with the same type of cancer as the test subject and a known prognosis (e.g., good or poor), immunotherapy outcome, or predicted survival time (e.g., less than 2 years, or 2 years or more).

[0073] In some embodiments, the methods can be used to monitor the efficacy of a treatment, e.g., an immunotherapy plus an NF-kB activator. The methods include determining levels of the NF-kB genes in a sample, then administering one or more doses of the treatment, then determining levels of the NF-kB genes to determine whether the treatment has increase immune infiltration of the tumor. An increase in NF-kB gene levels (or immune-related gene score, if calculated) indicates that the treatment was effective.

[0074] Immunotherapy

[0075] T cell presence in tumors is typically associated with immune surveillance and improved patient survival (Zhang et al. 2003. The New England Journal of Medicine 348:203-213; Fridman et al., 2011. Cancer research 71:5601-5605; Pages et al. 2005. The New England journal of medicine 353:2654-2666; Yu et al., 2009. Nature reviews. Cancer 9:798-809; Yu et al., 2007. Nature Reviews. Immunology 7:41-51; Schreiber et al., 2011. Science 331:1565-1570; Vesely et al., 2011. Annual Review of Immunology 29:235-271). Consequently, immunotherapy using blockade of negative regulators of T cells function is an especially attractive approach (Hodi et al., The New England journal of medicine 2010 363:711-723; Topalian et al., 2012. The New England Journal of Medicine 366:2443-2454). Unlike genetic lesion-specific therapies, immunotherapy has potential for targeting tumors irrespective of driver oncogene mutation status. Thus in some embodiments, the methods include administering an immunotherapy to the subject, e.g., one or more therapies that promote anti-cancer immunity, including administering one or more of: dendritic cells or peptides with adjuvant, immune checkpoint inhibitors, DNA-based vaccines, cytokines (e.g., IL-2), cyclophosphamide, agonists of OX40 (OX40; CD134), anti-interleukin-2R immunotoxins, and/or antibodies such as anti-CD137, anti-PD1, or anti-CTLA-4; see, e.g., Kruger et al., Histol Histopathol. 2007 Jun.; 22(6):687-96; Eggermont et al., Semin Oncol. 2010 October; 37(5):455-9; Klinke D J 2nd, Mol Cancer. 2010 Sep. 15; 9:242; Alexandrescu et al., J Immunother. 2010 July-August; 33(6):570-90; Moschella et al., Ann N Y Acad Sci. 2010 April; 1194:169-78; Ganesan and Bakhshi, Natl Med J India. 2010 Jan.-Feb.; 23(1):21-7; Golovina and Vonderheide, Cancer J. 2010 Jul.-Aug.; 16(4):342-7; Hodi et al., The New England journal of medicine 2010 363:711-723; Pentcheva-Hoang et al., Immunological Reviews 2009 229:67-87; Brahmer et al., Journal of Clinical Oncology 2010 28:3167-3175; Lynch et al., Journal of Clinical Oncology 2012 30(17):2046; Weber, Current Opinion in Oncology 2011 23:163-169; Weber, Seminars in Oncology 2010 37:430-439; Topalian et al., 2012. The New England Journal of Medicine 366:2443-2454; and Higano et al., Cancer 2009 115:3670-3679. In some embodiments, the methods include administering a composition comprising tumor-pulsed dendritic cells, e.g., as described in WO2009/114547 and references cited therein. Additional examples of immunotherapies include virus-based anti-cancer vaccines (e.g., adenovirus), formulations of Toll-like Receptor or RIG-I-like receptor ligands, Adoptive T cell therapy or other cell types. In some embodiments the immunotherapy is selected from the group consisting of BiovaxID (an autologous vaccine containing tumor-specific idiotype proteins from individual patient's lymphoma cells conjugated to keyhole limpet hemocyanin (KLH)); Provenge sipuleucel-T (an FDA-approved example of the use of autologous dendritic cells); Yervoy (a mAb against CTLA-4 (CD152), approved in 2011 for metastatic melanoma); tremelimumab (formerly ticilimumab, an anti-CTLA-4 mAb); IMA901 (a vaccine containing 10 tumor-associated peptides (TUMAPs)), alone or in combination with Sutent (a small molecule VEGF receptor tyrosine kinase inhibitor); GV1001 (a peptide vaccine with the sequence of human telomerase reverse transcriptase (hTERT), from Kael-Gemvax); Lucanix belagenpumatecel-L (four NSCLC cell lines carrying antisense oligonucleotides against transforming growth factor beta 2 (TGFB2)); Stimuvax (a liposomal vaccine containing a synthetic 25-amino acid peptide sequence from mucin 1 (MUC1; CD227)); Allovectin velimogene aliplasmid (a DNA plasmid encoding major histocompatibility complex (MHC) class I B7 (HLA-B7) complexed with lipid); BMS-936558 (ONO-4538) (a human mAb against PD-1); BMS-936559 (formerly MDX-1105) (a human mAb against PD-L1); Zelboraf (vemurafenib, an oral small molecule inhibitor of the oncogenic BRAF V600E mutation); Votrient (pazopanib, a small molecule VEGF receptor tyrosine kinase inhibitor); ISF35 or Lucatumumab (HCD122) (mAbs against CD40); GVAX (an allogeneic cancer vaccine engineered to secrete granulocyte macrophage-colony stimulating factor (GM-CSF)). See, e.g., Flanagan, "Immune Springboard," Biocentury, Jun. 18, 2012 A5-A10 (2012), available at biocentury.com. In some embodiments, the immunotherapy comprises administration of an agent that effects CTLA4 blockade (e.g., Ipilumumab BMS), PD1-blockade (e.g., BMS-936558, BMS; CT-011, Curetech; MK-3475, Merck), CD137 activation (e.g., BMS-663513, BMS), PD-L1 blockade (e.g., BMS-936559, BMS), CD40 activation (e.g., CP-870893, Pfizer) and autologous dendritic cells (e.g., Provenge).

[0076] NF-kB Inhibitors

[0077] In some embodiments, the methods include administering a therapy comprising an NF-kB inhibitor, i.e., a compound that inhibits the Nuclear Factor kappa B (NF-kB) intracellular transcription factor, to a subject who has a high level of NF-kB activity. Exemplary NF-kB inhibitors include sulfasalazine, Luteolin, rapamycin or derivatives (e.g., temsirolimus and everolimus), caffeic acid phenethylester, SN50 (a cell-permeable inhibitory peptide), parthenolide, triptolide, wedelolactone, lactacystin, substituted resorcinols, (E)-3-(4-methylphenylsulfonyl)-2-propenenitrile (e.g., Bay 11-7082, Bay 11-7821, or Bay 11-7085, Sigma-Aldrich, St. Louis, Mo.), Pranlukast, etoposide, bortezomib, MLN9708 (Kupperman et al., MLN9708), PS-1145 (Millennium Pharmaceuticals), tetrahydrocurcuminoids (such as Tetrahydrocurcuminoid CG, available from Sabinsa Corporation of Piscataway, N.J.), extracts of Paulownia tomentosa wood, and MG-132 [Z-Leu-Leu-Leu-H]. See, e.g. U.S. Pat. No. 7,838,513. Inhibitory nucleic acids targeting NF-kB, e.g., siRNA, antisense, or locked nucleic acids, can also be used. Hsp90 inhibitors such as 17-DMAG, AUY-922 and IPI-504 can also be used to inhibit NF-kB activation. See, e.g., US20070110828.

[0078] NF-kB Activators

[0079] In some embodiments, the methods include administering a therapy comprising an NF-kB activator, i.e., a compound that enhances activity of the Nuclear Factor kappa B (NF-kB) intracellular transcription factor, to a subject who has a low level of NF-kB activity, to enhance the anti-tumor immune response, e.g., in combination with immunotherapy. Exemplary NF-kB activators include cytotoxic anticancer agents such as taxanes (e.g., paclitaxel), vinca alkaloids (e.g., vinrelobine, vinblastine, vindesine, and vincristine), anthracyclines (e.g., daunorubicin, doxorubicin, mitoxantraone, and bisanthrene), epipodophyllotoxins (e.g., etoposide, etoposide orthoquinone, and teniposide); histone deacetylase inhibitors (e.g., romidepsin); pemetrexed; and topoisomerase inhibitors. See, e.g., Das and White, 1997. The Journal of Biological Chemistry, 272, 14914-14920; Nakanishi and Toi, Nat Rev Cancer. 2005 Apr.; 5(4):297-309; Ganapathi et al., Curr Pharm Des. 2002; 8(22):1945-58; see also Goodman and Gilman's The Pharmacological Basis of Therapeutics, e.g., 1277-1290 (7th ed. 1985) for descriptions and exemplary compounds.

EXAMPLES

[0080] The invention is further described in the following examples, which do not limit the scope of the invention described in the claims.

Example 1

Critical Role of NF-kB in Immunogenic Tumor Rejection in Mice

[0081] The presence of tumor infiltrating T cells, which likely recognize tumor expressed antigens, is associated with improved patient survival (Zhang et al. 2003. The New England Journal of Medicine 348:203-213; Fridman 2011. Cancer Research 71:5601-5605; Pages et al., 2005. The New England Journal of Medicine 353:2654-2666). To induce de novo anti-tumor T cell responses in mice, Kb-OVA (a single polypeptide encoding H-2K.sup.b, .beta..sub.2-M and the ovalbumin SIINFEKL peptide recognized by CD8 T cells)(Wang et al., 2009. Science 326:871-874) was expressed in poorly immunogenic Lewis lung carcinoma (LLC) to generate LLC-OVA.

[0082] The LLC cells were cultured in DMEM supplemented with 10% fetal bovine serum (FBS). Retroviruses were prepared by transfecting HEK 293T cells with Kb-OVA and packaging vectors as previously described (Valenzuela et al., 2009. J Clin Invest 119:3774-3786). Retrovirus transduced cells were sorted based on GFP expression using a FACS Vantage sorter (BD Biosciences, San Jose, Calif.) (Wang et al., 2007. J Immunol 178:6777-6788). Retrovirus infected LLC and human cell-lines were sorted based on GFP expression to yield >95% purity. Tetramer staining was performed as described (Cho and Celis, 2009. Cancer research 69:9012-9019) with the following changes: cells were incubated for 5 minutes at RT with Fc block and DAPI was added to cells prior to analysis for viability gating. Her/Neu tetramer has been described (Nava-Parada et al., 2007. Cancer research 67:1326-1334) and H2-Kb OVA tetramer was purchased from Beckman Coulter (Brea, Calif.). Flow cytometric analysis was performed on an LSR II cytometer (BD Biosciences, San Jose, Calif.). Aggregates and dead cells were excluded from analysis. Data were acquired using CellQuest software (BD Biosciences, San Jose, Calif.) and analyzed using FlowJo software (Tree Star, Ashland, Oreg.).

[0083] Cells were harvested in logarithmic growth after being cultured for less than two weeks and washed once in injection medium (phenol-free DMEM supplemented with 2% FBS) and counted. 5.times.10.sup.5 LLC cells were injected either s.c. (in a volume of 100 ul) or i.v. (in a volume of 200 ul). Subcutaneous tumors were monitored for growth and measured 2-3 times per week. Mice receiving intravenous LLC injections were monitored for morbidity. Mice were sacrificed when s.c. tumors reached a diameter of 20 mm or when they showed signs of morbidity (i.v. or s.c.). Tumor volume was calculated as previously described (Torabian et al. 2009. The American Journal of Pathology 174:1009-1016). Relative tumor growth between treatment groups was analyzed using the Student's T test with Welch's correction. Mice were maintained under specific pathogen free conditions.

[0084] Subcutaneous (s.c.) inoculation with LLC-OVA was sufficient to induce an OVA-specific CD8 T cell response (FIG. 1I).

[0085] The next experiments determined how tumor NF-kB activity impacts anti-tumor CD8 T cell responses. To selectively activate NF-kB in tumor cells, tumor cell-specific expression of constitutively-activated (CA)-IKK.beta. was utilized. IKK.beta. mediates NF-kB activation in response to multiple stimuli and pathways, including those activated by oncogenes such as KRAS (Basseres et al., 2010. Cancer Res 70:3537-3546). Furthermore, IKK.beta. is potentially amplified in human cancer (Beroukhim et al., 2010. Nature 463:899-905). Retroviruses were prepared by transfecting HEK 293T cells with MiG or activated IKK.beta. (S177, S181 to E mutations; IKK.beta.EE) and packaging vectors as previously described (Valenzuela et al., 2009. J Clin Invest 119:3774-3786; Wang et al., 2007. J Immunol 178:6777-6788).

[0086] CA-IKK.beta. expression in LLC (LLC-IKK) led to increased NF-kB, but not AP-1, nuclear translocation (FIG. 1J). Similarly, CA-IKK.beta. expression in LLC-OVA (LLC-OVA-IKK) enhanced nuclear NF-kB, comprising primarily of RelA/p65-containing complexes, and target gene CXCL1/KC expression (FIGS. 1A-B). However, both were induced substantially less compared to TNF.alpha. treatment (FIGS. 1A-B). Therefore, CA-IKK.beta. induces modest activation of NF-kB relative to TNF.alpha..

[0087] Compared to control LLC-MiG, LLC-IKK had no significant effect on s.c. growth of non-immunogenic LLC in syngeneic C57BL/6 mice (FIG. 1C). Interestingly, LLC-OVA-IKK tumors initially grew but were subsequently rejected while LLC-OVA-MiG grew unrestrained (FIG. 1D). Importantly, a similar number of activated OVA-specific CD8 T cells were detected in peripheral blood of LLC-OVA-MiG and LLC-OVA-IKK mice (FIG. 1E) suggesting that reduced growth of LLC-OVA-IKK was not due to impaired T cell priming Combined results from 3 experiments indicated that once tumors were perceptible (day 4), 10/11 LLC-OVA-MiG showed 2-fold or greater tumor growth while only 3/11 LLC-OVA-IKK showed similar growth (FIG. 1F). The difference in tumor numbers showing growth in the two groups was significant (p=0.008, Fisher's Exact Test). Importantly, LLC-OVA-IKK grew robustly in Rag2-/- mice (FIG. 1G), demonstrating a role for lymphocytes in rejection of LLC-OVA-IKK.

[0088] To extend these studies to a different lung tumor model, KRAS mutant LKR-13 cells (DuPage et al., 2011. Cancer cell 19:72-85) were used. As in LLC, CA-IKK.beta. expression in LKR-13 also resulted in NF-kB activation (FIG. 1L). Importantly, while LKR-OVA-MiG showed robust growth, LKR-OVA-IKK tumor growth was drastically reduced (FIG. 1H). These results therefore indicate that NF-kB activation induces rejection and/or growth suppression of immunogenic lung tumors in mice.

[0089] The next experiments used a metastatic model in which i.v. injected LLC form tumor foci in lungs. Importantly, LLC-OVA-IKK showed virtually no tumor foci compared to LLC-OVA-MiG, while both non-immunogenic LLC-MiG and LLC-IKK showing multiple foci (FIG. 2A). While few tumor cells were evident in lungs of LLC-OVA-IKK, multiple lymphoid aggregates were detected.

[0090] The effect of NF-kB activation was then determined in vaccine-induced responses against the breast carcinoma TUBO line, which expresses HER2/neu (Lee et al., 2010. Cancer immunology, immunotherapy: CII 59:1073-1081). Turin-Bologna (TUBO) cell-lines were cultured in DMEM supplemented with 10% fetal bovine serum (FBS), and transduced with the vectors as described above. TUBO-MiG and TUBO-IKK injected mice were randomly split into control (no vaccine) and TriVax (Assudani et al., 2008. Cancer Res. 68:9892-9899; Cho and Celis. 2009. Cancer Res. 69:9012-9019) vaccine groups, which were immunized using a synthetic peptide from HER2/neu (Nava-Parada et al., 2007. Cancer Res. 67:1326-1334). Mice receiving TUBO cells were split into TriVax treatment and non-treatment groups on day 5 Immunization with HER2 TriVax was performed as described (Nava-Parada et al., 2007. Cancer Res. 67:1326-1334).

[0091] TriVax-induced HER2-specific T cell increase in peripheral blood was similar in the two groups (FIG. 2C). Tumor growth of vaccinated TUBO-MiG and TUBO-IKK was then determined relative to their unvaccinated counterparts. While vaccination reduced growth of both TUBO-MiG and TUBO-IKK tumors, the reduction was significantly more pronounced in TUBO-IKK tumors (p=0.025) (FIG. 2B). Thus, NF-kB activation restrains tumor growth in de novo and vaccine-induced T cell models.

Example 2

NF-kB Induced T Cell Chemokine Expression is Crucial for Tumor Rejection

[0092] Possible mechanisms involved in rejection of LLC-OVA-IKK tumors were investigated. Rejection was not likely mediated by increased numbers of OVA-specific CD8 T cells (FIG. 1E). In addition, LLC-OVA-IKK cells were not superior stimulators of CD8 T cell IFN.gamma. expression (FIG. 3A). Importantly, while LLC-OVA-MiG tumors had a small number of infiltrating CD8 T cells, LLC-OVA-IKK tumors showed greatly increased CD8 T cell presence. These results suggested that NF-kB regulated expression of T cell chemokines may be responsible for increased T cell recruitment.

[0093] To identify T cell chemokines involved, global RNA expression studies using microarray analysis was performed on LLC-OVA-MiG and LLC-OVA-IKK. RNA was isolated using a Qiagen RNeasy kit, then reverse-transcribed and subjected to quantitative PCR analysis as described (Wang et al., 2007. J Immunol 178:6770-6776) in an Applied Biosystems 7900HT Sequence Detection System (Carlsbad, Calif.) with SYBR Green I Master Mix (Applied Biosystems, Carlsbad, Calif.) using gene-specific primers. All samples were run in triplicate and were normalized to rRNA 18s or .beta.-actin. Primers were obtained from RealTimePrimers.com. For microarray analysis the mRNA in 100 ng of total RNA was specifically converted to cDNA and then amplified and labeled with biotin using the Ambion Message Amp Premier RNA Amplification Kit (Life Technologies, Grand Island, N.Y.) following the manufacturer's protocol. Affymetrix Mouse Genome 430 2.0 Arrays were used in these studies. Hybridization with the biotin-labeled RNA, staining, and scanning of the arrays following the prescribed procedure outlined in the Affymetrix technical manual. Results were analyzed using the MAS 5.0 algorithm. Genes were considered changed if they were identified as changed in the MAS 5.0 comparison analysis and there was a 2-fold difference in signal compared to the control condition.

[0094] A 2-fold cutoff was used to identify genes up-regulated or down-regulated by IKK.beta. in two independent experiments. In total, 88 genes were up-regulated and 83 genes were down-regulated in both experiments. These included multiple chemokines involved in both T cell and neutrophil chemotaxis (FIG. 3B and FIG. 3E). Amongst T cell chemokines identified, CCL2, CCL5 and CXCL10 are known to mediate activated T cell chemotaxis (Bromley et al., 2008. Nature immunology 9:970-980). RT-PCR confirmed upregulation of CCL2 and CCL5 in LLC-OVA-IKK (FIG. 3C). Similarly, CCL2 and CCL5 expression was also substantially enhanced in LKR-OVA-IKK (data not shown). shRNA mediated knock-down (KD) showed that RelA plays a more important role than cRel or RelB in expression of T cell chemokines in LLC-OVA-IKK.

[0095] CCL2 in particular exhibited a dramatically higher microarray probe set signal than T cell chemokines CCL5 and CXCL10 (FIG. 3B) suggesting it may dominate T cell recruitment responses by LLC-OVA-IKK. In addition, CCL2 protein expression was greatly increased in LLC-OVA-IKK compared to LLC-OVA-MiG (FIG. 3F), as determined by ELISA using the following methods. Supernatant was collected from LLC-OVA-MiG cells and LLC-OVA-IKK cells after 24 h of culture. Anti-CCL2 ELISA was performed to measure CCL2 production in cells and to confirm CCL2 knockdown using the Mouse CCL2 ELISA Ready-SET-Go.RTM. kit from eBioscience (San Diego, Calif.) according to the manufacturer's instructions.

[0096] Previous studies have indicated an important role for CCL2 in breast cancer metastasis through monocyte recruitment (Qian et al., 2011. Nature 475:222-225) but also in T cell recruitment (Harlin et al., 2009. Cancer research 69:3077-3085). Given high expression and potentially diverse functions, it was determined whether CCL2 was specifically required for rejection of LLC-OVA-IKK tumors by KD of CCL2 expression. CCL2-KD did not impact LLC-OVA-IKK growth in vitro or OVA-specific T cell expansion in vivo. However, while control LLC-OVA-IKK tumors were readily rejected, CCL2-KD resulted in robust tumor growth (FIG. 3D). These results therefore suggest that CCL2 is a potentially crucial immune surveillance regulating NF-kB target gene that is required for LLC-OVA-IKK rejection.

Example 3

Generation of a Gene Expression Signature to Predict NF-kB Activity in Human Lung Cancer

[0097] Given the above findings in mice, it was next determined whether NF-kB activity is also associated with T cell presence in human lung cancer. It was hypothesized that an NF-kB-driven gene expression signature will provide a superior indication of NF-kB activity than activation state of individual subunits or NF-kB pathway kinases. Furthermore, previous studies have shown the predictive potential of gene expression signatures in determining pathway activation state (Bild et al. 2006. Nature 439:353-357; Chang et al. 2009. Mol Cell 34:104-114). To the best of the present inventors' knowledge, no such gene expression signature exists to predict NF-kB activity in human lung cancer.

[0098] Five human lung cancer cell-lines (A549, H23, H358, PC9 and HCC827) were used to generate such a signature. In each cell-line, genes were identified that were impacted by NF-kB inhibition (using retrovirus-expressed IkB.alpha.SR "super-repressor" of NF-kB) (Wang et al., 2007. J Immunol 178:6770-6776) and/or NF-kB activation (using CA-IKK.beta. described above) (Valenzuela et al. 2009. J Clin Invest 119:3774-3786).

[0099] As shown in FIG. 4A, IkB.alpha.SR reduced and CA-IKK.beta. increased NF-kB heterodimer nuclear activity compared to control MiG retrovirus infected cells. Although NF-kB is widely considered to be a tumor growth-promoting transcription factor, there was surprisingly little effect on survival or proliferation of these 5 lung cancer cell-lines following NF-kB inhibition by IkB.alpha.SR expression.

[0100] RNA isolated from these 5 cell-lines was used to perform gene expression studies using Affymetrix U133 Plus 2.0 microarrays. Initially, all probe sets impacted 1.4-fold or more by IkB.alpha. or IKK.beta. relative to the control vector were identified in each cell-line. This low stringency cutoff was used to accommodate a second selection criteria. Next, the 5 cell-line signatures were used to identify probe sets similarly regulated by IkB.alpha. and IKK.beta. (with IkB.alpha. and IKK.beta. causing opposite effects on gene expression) in the different cell lines in .gtoreq.60% of the experimental conditions (i.e. co-regulation score of .gtoreq.0.6). A total of 240 probe sets were identified using this approach (Table 1). The rationale for this was that the identification of genes similarly regulated by NF-kB in multiple cell lines will not only eliminate false positives but also generate a signature that is broadly applicable. As examples, 2 probe sets for the well-known NF-kB target gene BIRC3/cIAP2 (Chu et al., 1997. Proc Natl Acad Sci USA 94:10057-10062; Wang et al., 1998. Science 281:1680-1683) had co-regulation scores of 1.0 and 0.8, and both probe sets of another known NF-kB target gene TNFAIP3/A20 (Harhaj and Dixit, 2011. Cell research 21:22-39) had co-regulation scores of 1.0. In addition, IkB.alpha. increased expression of certain genes while IKK.beta. repressed their expression; these likely represent genes negatively impacted by NF-kB activity. In the 240 probe sets (Table 1), .about.200 individual genes were present that were either up-regulated or down-regulated by NF-kB.

TABLE-US-00002 TABLE 1 Up/Down Regulated Probe ID GenBank ID GeneSymbol Up 209099_x_at NM_000214.1 JAG1 Up 216268_s_at NM_000214.1 JAG1 Up 202076_at NM_001166.3 BIRC2 Up 210538_s_at NM_182962.1 BIRC3 Up 230499_at NM_001165 BIRC3 Up 225285_at NM_005504.4 BCAT1 Up 202357_s_at NM_001710.4 BF Up 217767_at NM_000064.1 C3 Up 226016_at NM_001025079.1 CD47 Up 211075_s_at NM_001025079.1 CD47 Up 213857_s_at NM_001025079.1 CD47 Up 227259_at NM_001025079.1 CD47 Up 202902_s_at NM_004079.3 CTSS Up 221903_s_at NM_015247 CYLD Up 203725_at NM_001924.2 GADD45A Up 224833_at NM_005238.2 ETS1 Up 1555355_a_at NM_005238.2 ETS1 Up 202269_x_at NM_002053.1 GBP1 Up 231577_s_at NM_002053.1 GBP1 Up 204224_s_at NM_001024070.1 GCH1 Up 223278_at NM_004004.3 GJB2 Up 204470_at NM_001511.1 CXCL1 Up 209774_x_at NM_002089.1 CXCL2 Up 207850_at NM_002090.2 CXCL3 Up 208729_x_at NM_005514.5 HLA-B Up 209140_x_at NM_005514.5 HLA-B Up 211911_x_at AY732487.1 HLA-C Up 202637_s_at NM_000201.1 ICAM1 Up 215485_s_at NM_000201.1 ICAM1 Up 202638_s_at NM_000201.1 ICAM1 Up 205207_at NM_000600.1 IL6 Up 204863_s_at NM_002184.2 IL6ST Up 211000_s_at NM_002184.2 IL6ST Up 212195_at NM_002184 IL6ST Up 212196_at NM_002184 IL6ST Up 202859_x_at NM_000584.2 IL8 Up 207536_s_at NM_001561.4 TNFRSF9 Up 211786_at NM_001561.4 TNFRSF9 Up 210511_s_at NM_002192.2 INHBA Up 231779_at NM_001570.3 IRAK2 Up 202531_at NM_002198.1 IRF1 Up 205032_at NM_002203.2 ITGA2 Up 227314_at NM_002203 ITGA2 Up 201389_at NM_002205.2 ITGA5 Up 205786_s_at NM_000632.3 ITGAM Up 202351_at NM_002210.2 ITGAV Up 202267_at NM_005562.1 LAMC2 Up 207517_at NM_018891.1 LAMC2 Up 207339_s_at NM_009588.1 LTB Up 203936_s_at NM_004994.2 MMP9 Up 209239_at NM_003998.2 NFKB1 Up 209636_at NM_002502.2 NFKB2 Up 207535_s_at NM_002502.2 NFKB2 Up 203927_at NM_004556.2 NFKBIE Up 1553994_at NM_002526 NT5E Up 201981_at NM_002581.3 PAPPA Up 206034_at NM_198833.1 SERPINB8 Up 214866_at NM_001005376.1 PLAUR Up 209040_s_at NM_004159.4 PSMB8 Up 204279_at NM_002800.4 PSMB9 Up 201762_s_at NM_002818.2 PSME2 Up 206157_at NM_002852.2 PTX3 Up 205205_at NM_006509.2 RELB Up 213194_at NM_002941.2 ROBO1 Up 216598_s_at NM_002982.3 CCL2 Up 1405_i_at NM_002985.2 CCL5 Up 205476_at NM_004591.1 CCL20 Up 202307_s_at NM_000593.5 TAP1 Up 225973_at NM_000544.3 TAP2 Up 204769_s_at NM_000544.3 TAP2 Up 208829_at NM_003190.3 TAPBP Up 204924_at NM_003264.3 TLR2 Up 202643_s_at NM_006290.2 TNFAIP3 Up 202644_s_at NM_006290.2 TNFAIP3 Up 205599_at NM_005658.3 TRAF1 Up 235116_at NM_005658.3 TRAF1 Up 208315_x_at NM_003300.2 TRAF3 Up 203299_s_at NM_003916 AP1S2 Up 222549_at NM_021101.3 CLDN1 Up 218182_s_at NM_021101.3 CLDN1 Up 203828_s_at NM_001012638.1 IL32 Up 204440_at NM_001040280.1 CD83 Up 221841_s_at NM_004235.3 KLF4 Up 212636_at NM 206853.1 QKI Up 204702_s_at NM_004289.5 NFE2L3 Up 236471_at NM_004289 NFE2L3 Up 204589_at NM_014840.2 NUAK1 Up 205100_at NM_005110.1 GFPT2 Up 1559064_at NM_005124 NUP153 Up 20320_at NM_005475.1 SH2B3 Up 229564_at NM_005493 RANBP9 Up 201489_at NM_005729.3 PPIF Up 202073_at NM_001008211.1 OPTN Up 202074_s_at NM_001008211.1 OPTN Up 207196_s_at NM_006058.2 TNIP1 Up 204341_at NM_006470.3 TRIM16 Up 213667_at NM_006662 SRCAP Up 220091_at NM_017585.2 SLC2A6 Up 227846_at BC067106.1 GPR176 Up 218644_at NM_016445.1 PLEK2 Up 213524_s_at NM_015714.2 G0S2 Up 220054_at NM_016584.2 IL23A Up 209788_s_at NM_001040458.1 ARTS-1 Up 218834_s_at NM_178031.2 HSPA5BP1 up 218627_at NM_018370.1 DRAM Up 219901_at NM_018351.2 FGD6 Up 220658_s_at NM_020183.3 ARNTL2 Up 224204_x_at NM_020183.3 ARNTL2 Up 222001_x_at XM_934503.1 FAM91A2 Up 229429_x_at XM_934505.1 FAM91A2 Up 209267_s_at NM_022154.5 SLC39A8 Up 219209_at NM_022168.2 IFIH1 Up 219759_at NM_022350.1 ERAP2 Up 218543_s_at NM_022750.2 PARP12 Up 222692_s_at NM_022763.2 FNDC3B Up 229865_at BC012204.1 FNDC3B Up 1568609_s_at NM_024615.2 PARP8 Up 219033_at NM_024615.2 PARP8 Up 220987_s_at NM_030952.1 NUAK2 Up 54970_at NM_031449.3 ZMIZ2 Up 223484_at NM_032413.2 C15orf48 Up 204823_at NM_014903.3 NAV3 Up 232593_at NM_138397 LOC93082 Up 220975_s_at NM_030968.2 C1QTNF1 Up 240770_at NM_173490 LOC134285 Up 205673_s_at NM_001031739.1 ASB9 Up 227599_at NM_178496.2 C3orf59 Up 226725_at NM_144975 MGC19764 Up 241379_at NM_173545.1 C2orf13 Up 212989_at NM_147156.3 TMEM23 Up 225105_at NM_207376.1 OCC-1 Up 229872_s_at XM_498811.2 KIAA0493 Up 229264_at XM_930678.1 LOC642441 Up 1555673_at XM_937100.1 LOC728285 Up 202660_at NM_002223.2 ITPR2 Up 227140_at NM_002192.2 INHBA Up 232060_at AK000776 ROR1 Up 232797_at AU144005 ITGAV Up 230787_at AW197616 Up 236704_at BG413366 Down 202053_s_at NM_000382.2 ALDH3A2 Down 202054_s_at NM_000382.2 ALDH3A2 Down 209442_x_at NM_020987.2 ANK3 Down 231729_s_at NM_004058.2 CAPS Down 231728_at NM_004058.2 CAPS Down 211922_s_at NM_001752.2 CAT Down 204920_at NM_001875.2 CPS1 Down 217564_s_at NM_001875.2 CPS1 Down 202295_s_at NM_004390.2 CTSH Down 209782_s_at NM_001352.2 DBP Down 215116_s_at NM_001005336.1 DNM1 Down 37996_s_at NM_004409.2 DMPK Down 204540_at NM_001958.2 EEF1A2 Down 36499_at NM_001408.1 CELSR2 Down 212339_at NM_177996.1 EPB41L1 Down 209189_at NM_005252.2 FOS Down 224863_at NM_002072 GNAQ Down 202756_s_at NM_002081.1 GPC1 Down 1555037_a_at NM_005896.2 IDH1 Down 201193_at NM_005896.2 IDH1 Down 210046_s_at NM_002168.2 IDH2 Down 219922_s_at NM_021070.2 LTBP3 Down 225540_at NM_002374.3 MAP2 Down 209035_at NM_001012333.1 MDK Down 203238_s_at NM_000435.1 NOTCH3 Down 212151_at NM_002585 PBX1 Down 204476_s_at NM_000920 PC Down 206348_s_at NM_005391.1 PDK3 Down 203407_at NM_002705.3 PPL Down 213093_at NM_002737.2 PRKCA Down 207011_s_at NM_152880.2 PTK7 Down 203453_at NM_001038.4 SCNN1A Down 208998_at NM_003355.2 UCP2 Down 208997_s_at NM_003355.2 UCP2 Down 205538_at NM_003389 CORO2A Down 211712_s_at NM_003568.1 ANXA9 Down 209641_s_at NM_003786.2 ABCC3 Down 202481_at NM_004753.4 DHRS3 Down 223761_at NM_005117.2 FGF19 Down 225112_at NM_005759 ABI-2 Down 1554018_at NM_002510 GPNMB Down 203961_at NM_006393 NEBL Down 209791_at NM_007365 PADI2 Down 212325_at NM_014988.1 LIMCH1 Down 212328_at NM_014988.1 LIMCH1 Down 212327_at BX537916.1 LIMCH1 Down 227031_at NM_015132 SNX13 Down 202341_s_at NM_015271.2 TRIM2 Down 202342_s_at NM_015271.2 TRIM2 Down 205499_at NM_014467.1 SRPX2 Down 219188_s_at NM_014067.2 MACROD1 Down 221922_at NM_013296.3 GPSM2 Down 219014_at NM_016619.1 PLAC8 Down 220779_at NM_016233.1 PADI3 Down 225355_at AK096661.1 DKFZP761M1511 Down 222496_s_at NM_019027.1 RBM47 Down 221748_s_at AF172820.1 MXRA6 Down 232914_s_at NM_032379.3 SYTL2 Down 218736_s_at NM_017734.2 PALMD Down 226050_at NM_017905.3 TMCO3 Down 218729_at NM_020169.2 LXN Down 220246_at NM_020397.2 CAMK1D Down 213839_at AB007969.1 CLMN Down 232116_at NM_021180.2 GRHL3 Down 218858_at NM_022783.1 DEPDC6 Down 219263_at NM_024539.3 RNF128 Down 218342_s_at NM_024896.2 KIAA1815 Down 222603_at NM_024896.2 KIAA1815 Down 220584_at NM_025094 FLJ22184 Down 224150_s_at NM_024491.2 CEP70 Down 238154_at NM_024491 BITE Down 228084_at NM_030821 PLA2G12 Down 212936_at NM_032042.3 C5orf21 Down 232176_at NM_032229.2 SLITRK6 Down 235976_at NM_032229.2 SLITRK6 Down 232481_s_at NM_032229.2 SLITRK6 Down 227134_at NM_032872.1 SYTL1 Down 228072_at NM_177963.2 SYT12 Down 226597_at NM_138393.1 C19orf32 Down 208158_s_at NM_018030.3 OSBPL1A Down 209485_s_at NM_080597.2 OSBPL1A Down 226325_at NM_199165.1 ADSSL1 Down 225240_s_at NM_138962 MSI2 Down 234974_at NM_138801 LOC130589 Down 235256_s_at NM_138801.1 GALM Down 226875_at NM_144658.2 DOCK11 Down 238029_s_at NM_152527.3 SLC16A14 Down 244261_at NM_170743.2 IL28RA Down 229616_s_at NM_001012642.1 GRAMD2 Down 224990_at NM_174921.1 C4orf34 Down 227052_at NM_174921 LOC201895 Down 228067_at NM_207362.1 C2orf55 Down 224989_at AI824013 C4orf34 Down 235924_at N73742 Down 1565837_at AA215492 Down 227533_at AA732944 RALGPS2 Down 228959_at AI676241 Down 232656_at AU145501 NAALADL2 Down 238441_at NM_006252.3 PRKAA2 Down 238105_x_at AW294903 WNT7B

[0101] A substantial number of NF-kB signature genes regulate immune cell chemotaxis (including T cell chemokines CCL2 and CCL5), inflammation and immune regulation (Table 2).

TABLE-US-00003 TABLE 2 NF-.kappa.B signature genes involved in inflammation and immune regulation Inflammation Adaptive Adhesion Complement IL-6 HLA-B CD47 CFB IL-6 ST HLA-C ITGA2 C3 IL-8 ICAM1 ITGA5 IRAK2 TNFRSF9 ITGAM IRF1 LTB ITGAV CXCL1 CCL2 LAMC2 CXCL2 CCL5 CLDN1 CXCL3 TAP1 PLAUR TAP2 CCL20 TAPBP TLR2 CD83 TRAF1 IL23A TRAF3 PSMB8 IFIH1 PSMB9 PSME2 IL32 NF-kB signature genes divided in broad categories based on known functions. The category distinction is not absolute since many genes are involved in multiple listed or other functions. The "Inflammation" category refers to genes involved in inflammation and innate immunity. "Adaptive" refers to genes involved in the adaptive immune response. "Adhesion" refers to genes involved in cell-cell or substrate interactions. "Complement" includes genes involved in the complement pathway.

[0102] Using publically available microarray data from 126 different human lung cancer cell lines, cell-lines with high or low NF-kB signature activity were next identified. The microarray data was obtained from GEO at NCBI and Array Express at EBI. It consists of 408 Affymetrix CEL files of 126 different lung cell lines. The scores were calculated based on this set of samples so the high and low are in reference to other cell lines within this group. The probesets that were used to classify the samples were the 240 probes originally identified in the 5 cell lines (i.e. NF-kB signature). The classifier was built and implemented as described (classifier H in supplemental materials and methods) (Shedden et al. 2008. Nat Med 14:822-827). The decision thresholds were made based on the array of measures in the 408 lung CEL file data. The weighted voting classification of each sample scores each gene (probeset) based on where the signal value falls among all the samples in the group. If the probeset was positively correlated with the NF-KB signature (in the original 5 cell lines) then values in the lower third receive a score of -1, values in the middle third receive a 0, and values in the upper third receive a value of 1. If a gene was negatively correlated the values were reversed. The scores for all probesets were summed to provide the final classification score. The data sets used are listed below:

TABLE-US-00004 E-MTAB-37 Array Express GSE10021 GEO GSE10843 GEO GSE13309 GEO GSE14315 GEO GSE14883 GEO GSE15240 GEO GSE16194 GEO GSE17347 GEO GSE18454 GEO GSE21612 GEO GSE4824 GEO GSE5816 GEO GSE6013 GEO GSE7562 GEO GSE8332 GEO

NF-kB signature scores were determined by building a classifier that allowed determination of relative signature activity in the different cell-lines, to classify cell lines with low or high NF-kB signature. The classifier was built using methods previously described (Shedden et al. 2008. Nat Med 14:822-827), as follows: A majority vote classifier was used based on the probesets identified in the NF-kB signature. The algorithm consists of three components: the majority vote classifier, the individual classifiers, and the training procedure. The classifier is a majority vote classifier based on the probesets identified that correlate with NF-kB activity in cell lines. The algorithm consisted of three components: the majority vote classifier, the individual classifiers, and the training procedure.

[0103] Terminology

[0104] C=majority vote of individual classifiers (sum of vote by all probesets in classifier)

[0105] c.sub.k=individual classifier k (a single probeset and thresholds for that probeset)

[0106] x.sub.j=sample j (microarray dataset for an array from a single tumor sample)

[0107] G=set of genes used for individual classifiers (set of all probesets in classifier)

[0108] g.sub.jk=gene expression value for sample j and probeset k

[0109] S=sign (+/-) indicating trend relative to NF-kB activity, +=high expression when NF-kB active=Blue group, -=high expression when NF-kB inactive=Rosy group.

[0110] Majority Vote Classifier

[0111] Cj=.SIGMA. (c.sub.kj) was the continuous predictor for this classifier, which was categorized into three groups according to the following rule:

Pred ( xj ) = { Blue C j > 0.15 G Rosy C j < - 0.15 G Grey otherwise ##EQU00001##

[0112] The value of 0.15 was heuristically determined depending on the specific prediction corresponding to the Blue or Rosy class. A Pred value of |G| (or -1*|G|) indicated complete agreement with Rosy or Blue group while small values (e.g., 0.15*|G|) indicated uncertainty in voting. As the NF-kB signature may predict NF-kB activity, response to chemotherapy, prognosis, infiltration of lymphocytes, and other biological properties the training process can be used to further tune the constant to maximize predictive success.

[0113] Individual Classifier

[0114] For each gk element of G [0115] Establish 3 quantiles defined by 2 threshold values, LO and HI:

[0115] ck = { - 1 * S k g jk < LO + 1 * S k g jk > HI 0 otherwise ##EQU00002##

[0116] This had the effect of voting -1 for individual genes under-expressed in the samples and +1 for over-expressed genes. The vote was reversed if the gene was negatively correlated to the two groups.

[0117] Training Process

[0118] The training process required a gene set G (e.g., a gene set identified elsewhere in this proposal) and a reference dataset for each sample class (tumor group) and a biologically known property that might be predicted based on NF-kB activity. The training process began by identifying the LO and HI thresholds for each element gk. All gene expression values were ranked in the reference data set from lowest expression level to highest expression level. The LO threshold was the 33rd percentile value for gene k within all samples of a given tumor class. This value could also be set using a reference dataset of the tumor class to be classified. Likewise the HI threshold was the 66th percentile value for gene k within all samples of a given tumor class or the reference dataset. Once the thresholds were set, the individual classification was applied to all genes (gk).

[0119] Next the individual classifiers were summed to generate the Majority Vote score Cj. This represented a continuous variable that represents the relative NF-kB activity in the assessed sample. Classification actually involved defining groups based on the separation of this continuous variable into discrete groups: blue and rosy. The constant (0.15 in the present example) was adjusted to maximally correlate the blue group or rosy group with the biological property that is predicted by the Majority Vote Classifier. This constant was set using the reference dataset for each tumor group and biological process.

[0120] Next, it was determined whether differences in NF-kB signature correlated with differences NF-kB nuclear presence, i.e., activity. Four NF-kB low and 4 NF-kB high signature cell lines were compared in a side by side analysis using EMSA. Based on this analysis, 4 cell lines that were identified with high NF-kB signature (H226, H157, H1299, H650) and 4 cell lines with low NF-kB signature (H322, H1395, H522, H1437) were used for additional studies.

[0121] Genes present in the NF-kB signature may also be regulated by non-NF-kB pathways. Thus, while H226, H157, H1299 and H650 cell-lines have high expression of NF-kB signature genes, it is unclear whether this is indeed due to NF-kB activity. To test this, expression of the two signature genes with high co-regulation scores (BIRC3 and TNFAIP3) was determined in these 8 cell-lines. Despite variation in expression in individual cell-lines, NF-kB high cell-lines had significantly higher mean expression of both BIRC3 and TNFAIP3 compared to NF-kB low cell-lines (FIG. 4B).

[0122] To determine NF-kB involvement, H226, H157 and H1299 cells were transduced with MiG and IkB.alpha.SR retroviruses (H650 showed little or no retrovirus infection). Importantly, expression of both genes was significantly reduced by IkB.alpha.SR expression (FIG. 4C). Similar results were obtained for additional genes including CCL2, CCL5, CXCL1-3 and IL8 (FIG. 4D). These results therefore indicate that high expression of NF-kB signature genes in these cell-lines is dependent on NF-kB activity.

[0123] The next experiments were performed to evaluate an association between the NF-kB signature and NF-kB DNA binding activity. Importantly, EMSA showed that the 4 lines with high NF-kB signature indeed had high kB-site binding activity (FIG. 4E). However, the NF-kB signature low H322 cell-line also showed high NF-kB activity (FIG. 4E). Therefore, NF-kB DNA binding activity alone does not provide an unambiguous indication of NF-kB transcriptional activity.

[0124] To seek a better understanding of NF-kB function in human lung cancer, the signature was next used for studies in human lung adenocarcinoma. Specifically, the correlation between NF-kB regulated chemokine gene expression and T cell presence was determined, as well as the association of these genes with patient survival.

Example 4

Association of T Cell Chemokines, but not Neutrophil Chemokines, with T Cell Presence in Human Lung Cancer

[0125] The association of different chemokines present in the NF-kB signature with T cell presence was determined. For these studies, the Consortium for the Molecular Classification of Lung Adenocarcinoma (CMCLA) survival prediction study (59) was used. CMCLA is the largest and most comprehensive study in which microarray-based gene expression data from tumors was used to predict survival of 442 early stage lung adenocarcinoma patients (Shedden et al. 2008. Nat Med 14:822-827). Tumor samples used were from four different institutions including the Moffitt Cancer Center (Shedden et al. 2008). Specifically, the experiments determined whether neutrophil (CXCL1-3 and IL8) and T cell chemokines (CCL2, CCL5 and CXCL10) were differentially associated with T cell presence (although CXCL10 did not achieve the criteria for inclusion in the NF-kB signature, it was included in this analysis because it was identified as a target gene in LLC and in some human cell-lines). To detect T cells, T cell receptor .alpha. (TRAC) and .beta. chain (TRBC1) gene expression was used as a marker for T cell presence. As expected, TRAC expression was highly correlated with TRBC1 expression (FIG. 5A). Importantly, CCL2, CCL5 and CXCL10, but not neutrophil chemokine expression (CXCL1-3 and IL8), was positively correlated with these T cell markers (FIG. 5A). These results therefore indicate a functional link between T cell chemokine expression and T cell presence in tumors. Interestingly, expression of neutrophil chemokine genes was highly correlated with each other (FIG. 5A) suggesting that NF-kB regulated inflammatory and immune response functions predominate in different tumors. To help understand how genes associated with different NF-kB functions may be differentially regulated, their association with key NF-kB activators was determined.

Example 5

LT.beta. Expression is Differentially Associated with T Cell Chemokines in Human Lung Cancer

[0126] Multiple NF-kB-activating cytokines may be present in the lung tumor microenvironment (Mantovani, A., Allavena, P., Sica, A., and Balkwill, F. 2008. Cancer-related inflammation. Nature 454:436-444). To identify cytokines potentially involved in differential expression of T cell versus neutrophil chemokines, association with expression of TNF.alpha., IL-1.alpha., IL-1.beta., LT.alpha.(TNF.beta.) and LT.beta. was tested. These cytokines not only induce NF-kB activation but can also be transcriptionally regulated by NF-kB. However, only LT.beta. was identified as an NF-kB target gene in lung cancer cells (Table 2) and interestingly, only LT.beta. showed a greater correlation with expression of T cell chemokines versus neutrophil chemokines (FIGS. 5B and 5E). LT.beta. receptor (LT.beta.R) engagement by heterodimers of LT.alpha. and LT.beta. (LT.alpha.1/.beta.2) activates NF-kB RelA and RelB, typically in association with p52 (Bonizzi and Karin, 2004. Trends Immunol 25:280-288; Dejardin et al., 2002. Immunity 17:525-535; Gommerman and Browning, 2003. Nature reviews. Immunology 3:642-655; Wolf et al., 2010. Oncogene 29:5006-5018). Mirroring correlation in tumors, soluble LT.alpha.1/.beta.2 (LT) induced expression of T cell chemokines but not neutrophil chemokines in human lung cancer cells (FIG. 5C). In contrast, TNF.alpha. strongly induced both neutrophil and T cell chemokines (FIG. 5D). These results suggest that differential expression of NF-kB regulated T cell chemokines versus neutrophil chemokines in human lung cancer could be achieved through agents that selectively induce these different chemokine subsets, such as the LT.beta.R ligand.

Example 6

NF-kB Signature Genes are Associated with Distinct Overall Survival of Lung Cancer Patients

[0127] The above results indicate that NF-kB regulated inflammatory and immune response genes are differentially expressed in lung cancer. The next experiments determined whether expression of these genes was also associated with distinct overall survival (OS) of patients. Using the CMCLA dataset, 5-year OS of patients exhibiting high versus low expression of these genes was determined using a median cutoff. A striking effect on survival was seen for NF-kB target genes involved in neutrophil chemotaxis and inflammation (IL8, CXCL1, CXCL3 and IL6), all of which were associated with significantly poor OS (FIG. 6A-D).

[0128] Poor survival in human cancer is associated with increased metastasis. Both IL-8 and CXCL1 have been linked to increased metastasis through effects on infiltrating myeloid cells (Sparmann and Bar-Sagi, 2004. Cancer Cell 6:447-458; Acharyya et al. 2012. Cell 150:165-178), suggesting association with poor survival can be through increased metastatic dissemination. On the other hand, high expression of a subset of genes involved in T cell chemotaxis and T cell responses, including CCL2, LT.beta., ICAM-1 and CD83, was associated with significantly improved OS (FIG. 7A-D). LT.beta. association with improved OS was especially pronounced, perhaps because LT.beta. can enhance expression of multiple genes involved in T cell responses. Importantly, high expression of TCR.alpha. and .beta. chain genes was also associated with significantly improved OS in lung cancer patients (FIG. 7E). Therefore, while both T cell presence and expression of NF-kB signature genes associated with T cell responses is associated with improved OS, the high expression of inflammatory genes is associated with poor OS. Hence, distinct functions of NF-kB in human lung cancer are associated with potentially different survival outcomes.

Example 7

NF-kB Activity in Human Lung Cancer is Strongly Associated with T Cell Presence

[0129] The above findings indicate differential expression of distinct NF-kB target genes in tumors. By allowing investigation of combined expression of signature genes, principal component analysis (PCA) can simplify evaluation of pathway-dependent gene expression activity and has been widely used to derive and validate gene signatures in various cancer studies (Chen et al., 2011. Journal of the National Cancer Institute 103:1859-1870; Chen et al., 2010. Breast cancer research and treatment 119:335-346; Chen et al., 2010. Breast cancer research and treatment 120:25-34). The first principal component (1.sup.st PC), (principal component analysis--PCA) was used, as it accounted for the largest variability in the data, to represent the overall expression level for NF-kB activation. That is,

NF-kB activation score=.SIGMA.w.sub.ix.sub.i,

a weighted average expression among the NF-kB activated genes, where x.sub.i represents gene i expression level, w.sub.i is the corresponding weight (loading coefficient) with .SIGMA.w.sub.i.sup.2=1, and the w.sub.i values maximize the variance of .SIGMA.w.sub.ix.sub.i.

[0130] This approach has been used to derive a malignancy pathway gene signature in a breast cancer study (Chen et al., 2010. Breast Cancer Res Treat 119:335-346). The Cox proportional hazards model was used to analyze the continuous NF-kB activation score to determine if it can predict survival.

[0131] Of the 240 NF-kB signature probe sets present in Affymetrix U133 Plus 2.0 microarrays, 159 were present (Table 3) in the CMCLA dataset which used Affymetrix 133A (Shedden et al., 2008).

TABLE-US-00005 TABLE 3 Probe ID Gene Symbol Probe ID Gene Symbol 209099_x_at JAG1 218644_at PLEK2 216268_s_at JAG1 213524_s_at G0S2 202076_at BIRC2 220054_at IL23A 210538_s_at BIRC3 209788_s_at ERAP1 202357_s_at NA 218834_s_at TMEM132A 217767_at C3 218627_at DRAM1 211075_s_at CD47 219901_at FGD6 213857_s_at CD47 220658_s_at ARNTL2 202902_s_at CTSS 222001_x_at LOC728855 221903_s_at CYLD 209267_s_at SLC39A8 203725_at GADD45A 219209_at IFIH1 202269_x_at GBP1 219759_at ERAP2 204224_s_at GCH1 218543_s_at PARP12 204470_at CXCL1 219033_at PARP8 209774_x_at CXCL2 220987_s_at NA 207850_at CXCL3 54970_at ZMIZ2 208729_x_at HLA-B 204823_at NAV3 209140_x_at HLA-B 220975_s_at C1QTNF1 211911_x_at HLA-B 205673_s_at ASB9 202637_s_at ICAM1 212989_at SGMS1 215485_s_at ICAMI 202660_at ITPR2 202638_s_at ICAM1 202053_s_at ALDH3A2 205207_at IL6 202054_s_at ALDH3A2 204863_s_at IL6ST 209442_x_at ANK3 211000_s_at IL6ST 211922_s_at CAT 212195_at IL6ST 204920_at CPS1 212196_at IL6ST 217564_s_at CPS1 202859_x_at IL8 202295_s_at CTSH 207536_s_at TNFRSF9 209782_s_at DBP 211786_at TNFRSF9 215116_s_at DNM1 210511_s_at INHBA 37996_s_at DMPK 202531_at IRF1 204540_at EEF1A2 205032_at ITGA2 36499_at CELSR2 201389_at ITGA5 212339_at EPB41L1 205786_s_at ITGAM 209189_at FOS 202351_at ITGAV 202756_s_at GPC1 202267_at LAMC2 201193_at IDH1 207517_at LAMC2 210046_s_at IDH2 207339_s_at LTB 219922_s_at LTBP3 203936_s_at MMP9 209035_at MDK 209239_at NFKB1 203238_s_at NOTCH3 209636_at NFKB2 212151_at PBX1 207535_s_at NFKB2 204476_s_at PC 203927_at NFKBIE 206348_s_at PDK3 201981_at PAPPA 203407_at PPL 206034_at SERPINB8 213093_at PRKCA 214866_at PLAUR 207011_s_at PTK7 209040_s_at PSMB8 203453_at SCNN1A 204279_at PSMB9 208998_at UCP2 201762_s_at PSME2 208997_s_at UCP2 206157_at PTX3 205538_at CORO2A 205205_at RELB 211712_s_at ANXA9 213194_at ROBO1 209641_s_at ABCC3 216598_s_at CCL2 202481_at DHRS3 1405_i_at CCL5 203961_at NEBL 205476_at CCL20 209791_at PADI2 202307_s_at TAP1 212325_at LIMCH1 204769_s_at TAP2 212328_at LIMCH1 208829_at TAPBP 212327_at LIMCH1 204924_at TLR2 202341_s_at TRIM2 202643_s_at TNFAIP3 202342_s_at TRIM2 202644_s_at TNFAIP3 205499_at SRPX2 205599_at TRAF1 219188_s_at MACROD1 208315_x_at TRAF3 221922_at GPSM2 203299_s_at AP1S2 219014_at PLAC8 218182_s_at CLDN1 220779_at PADI3 203828_s_at IL32 221748_s_at TNS1 204440_at CD83 218736_s_at PALMD 221841_s_at KLF4 218729_at LXN 212636_at QKI 220246_at CAMK1D 204702_s_at NFE2L3 213839_at CLMN 204589_at NUAK1 218858_at DEPDC6 205100_at GFPT2 219263_at RNF128 203320_at SH2B3 218342_s_at ERMP1 201489_at PPIF 220584_at FLJ22184 202073_at OPTN 212936_at FAM172A 202074_s_at OPTN 208158_s_at OSBPL1A 207196_s_at TNIP1 209485_s_at OSBPL1A 204341_at TRIM16 213667_at SRCAP 220091_at SLC2A6

[0132] NF-kB signature "driver" genes (i.e., the top 10 genes with highest PCA weight) were identified as GBP1, PSMB9, IRF1, TAP1, TNFAIP3, CCL5, PSMB8, IL32, SH2B3 and NFKBIE; all are upregulated by NF-kB (see Table 1). First principal component (PC1) is associated with the largest variance of the data (e.g., variance of NF-kB signature activity in the CMCLA dataset) followed by each succeeding PC. Using association with expression of driver genes, it was found that PC1 but not PC2 is associated with NF-kB activity (FIG. 8A). Using the NF-kB signature PC1, it was determined whether NF-kB activity was associated with T cell presence. For these studies, PC1 of T cell receptor .alpha. and .beta. chain gene expression (FIG. 8B, top panel) was used. Importantly, a strong association was seen between NF-kB activity and T cell presence (r=0.68) (FIG. 8C). Therefore, while NF-kB target genes are associated with different functions in tumors, the overall NF-kB activity as determined by PCA is strongly associated with T cell presence.

[0133] In addition to inflammation, NF-kB also exerts pro-tumor effects through cancer cell-intrinsic regulation of cell survival and proliferation (Barbie et al. 2009. Nature 462:108-112; Meylan et al., 2009. Nature 462:104-107; Maeda et al., 2005. Cell 121:977-990; Karin and Greten, 2005. Nat Rev Immunol 5:749-759; Greten et al., 2004. Cell 118:285-296; Takahashi et al., 2010. Cancer Cell 17:89-97; Basseres et al., 2010. Cancer Res 70:3537-3546). Previous studies described a breast and NSCLC malignancy-risk (MR) signature that is rich in proliferation and cell cycle genes (Chen et al., 2011. Journal of the National Cancer Institute 103:1859-1870; Chen et al., 2010. Breast cancer research and treatment 119:335-346; Chen et al., 2010. Breast cancer research and treatment 120:25-34). Using the MR signature PC1, it was next determined whether NF-kB activity was also associated with cancer cell proliferation. However, little correlation between the NF-kB and MR signatures was noticed (r=0.21) (FIG. 8D). Thus, NF-kB activity is associated with T cell presence and potential immune surveillance functions but not with cancer cell proliferation.

[0134] Whether expression of the 10 driver genes mentioned above (GBP1, PSMB9, IRF1, TAP1, TNFAIP3, CCL5, PSMB8, IL32, SH2B3 and NFKBIE) was associated with overall NF-kB activity and T cell presence was also determined Indeed, PC1 of these genes (FIG. 8B, bottom panel) was highly correlated with the NF-kB signature PC1 (r=0.92) (FIG. 8E). Therefore, this smaller 10-gene signature can be used in lieu of the NF-kB signature to determine NF-kB activity. Importantly, the 10-gene PC1 was also strongly correlated with the T cell PC1 (r=0.79) (FIG. 8F) but not with the MR signature PC1 (r=0.19) (FIG. 8G). Using an additional lung cancer microarray dataset (n=133) (Zhu et al., 2010. Journal of Clinical Oncology 28:4417-4424), there was a similarly high correlation of the 10-gene signature with T cell presence (r=0.8) but not with the MR signature (r=-0.16) (FIG. 8H).

[0135] In conclusion, the present findings from both mouse and human studies indicate that tumor NF-kB activity is strongly associated with T cell presence and immune surveillance in lung cancer.

Example 8

NF-kB Signature Classifies Multiple Tumor Types into High and Low for the NF-kB Signature, and Correlates with Presence of T Cells

[0136] A majority vote classifier as mentioned above can be used to classify different tumor types (including colon adenocarcinoma, renal carcinoma, ovarian cancer, kidney tumor, prostate tumors, and melanoma) into high and low for NF-kB signature, with similar results expected. High NF-kB signature corresponds to positive numbers and low activity corresponds to negative numbers. Each tumor class had individual samples with high, intermediate, and low NF-kB signatures. Exemplary data is shown in FIG. 9, in which PCA, performed as described above, was used to demonstrate a high degree of correlation between T cell presence and the 159 genes listed in Table 3 (r=0.84) or the 10-gene signature (r=0.95) in melanoma. This demonstrated that the classification procedure can be used in many tumor types, and also that tumors exhibit a continuum of values representing their NF-kB activity.

Other Embodiments

[0137] It is to be understood that while the invention has been described in conjunction with the detailed description thereof, the foregoing description is intended to illustrate and not limit the scope of the invention, which is defined by the scope of the appended claims. Other aspects, advantages, and modifications are within the scope of the following claims.

Claims

1. A method of predicting NF-kB transcriptional activity in a tumor, the method comprising: determining gene expression levels for guanylate binding protein 1, interferon-inducible (GBP1); proteasome (prosome, macropain) subunit, beta type, 9 (large multifunctional peptidase 2) (PSMB9); interferon regulatory factor 1 (IRF1); transporter 1, ATP-binding cassette, sub-family B (MDR/TAP) (TAP1); tumor necrosis factor, alpha-induced protein 3 (TNFAIP3); chemokine (C-C motif) ligand 5 (CCL5); proteasome (prosome, macropain) subunit, beta type, 8 (large multifunctional peptidase 7) (PSMB8); interleukin 32 (IL32); SH2B adaptor protein 3 (SH2B3); and nuclear factor of kappa light polypeptide gene enhancer in B-cells inhibitor, epsilon (NFKBIE) in a sample comprising cells from the tumor; and comparing the gene expression levels to reference levels; optionally assigning a score to the tumor based on the comparison of the gene expression levels in the tumor to the reference levels; wherein the presence of gene expression levels above the reference levels, or a score above a threshold score, indicates that the tumor has high levels of NF-kB activity, and the presence of gene expression levels below the reference levels, or a score below the threshold score indicates that the tumor has low levels of NF-kB activity.

2. The method of claim 1, further comprising identifying a subject as having a tumor having high levels of NF-kB activity, and selecting a treatment for the subject comprising administering an NF-kB inhibitor or immunotherapy.

3. The method of claim 1, further comprising identifying a subject as having a tumor having low levels of NF-kB activity, and selecting a treatment for the subject comprising administering an NF-kB activator and immunotherapy.

4. A method of selecting a treatment for a subject who has a tumor, the method comprising: determining gene expression levels for GBP1, PSMB9, IRF1, TAP1, TNFAIP3, CCL5, PSMB8, IL32, SH2B3 and NFKBIE in a sample comprising cells from the tumor; and comparing the gene expression levels to reference levels; optionally assigning a score to the tumor based on the comparison of the gene expression levels in the tumor to the reference levels; detecting the presence of gene expression levels above the reference levels, or a score above a threshold score, and selecting a treatment for the subject comprising administering an NF-kB inhibitor or immunotherapy; or detecting the presence of gene expression levels below the reference levels, or a score below a threshold score, and selecting a treatment for the subject comprising administering an NF-kB activator and immunotherapy.

5. The method of claim 4, further comprising administering the selected treatment to the subject.

6. A method of predicting outcome in a subject with a tumor, the method comprising: determining gene expression levels for GBP1, PSMB9, IRF1, TAP1, TNFAIP3, CCL5, PSMB8, IL32, SH2B3 and NFKBIE in a sample comprising cells from the tumor; and comparing the gene expression levels to reference levels; optionally assigning a score to the tumor based on the comparison of the gene expression levels in the tumor to the reference levels; wherein the presence of gene expression levels below the reference levels, or a score below a threshold score, indicate that the subject is likely to have a poor outcome as compared to a subject who has levels above the reference levels or threshold score, and the presence of gene expression levels above the reference levels, or a score above a threshold score, indicate that the subject is likely to have a better outcome as compared to a subject who has levels below the reference levels or threshold score.

7. (canceled)

8. (canceled)

9. (canceled)

10. The method of any of claim 1, wherein the tumor is an adenocarcinoma.

11. The method of claim 10, wherein the adenocarcinoma is lung adenocarcinoma or melanoma.

12. The method of claim 1, wherein determining gene expression levels comprises performing an assay to determine gene expression levels in the sample.

13. The method of claim 1, wherein assigning a score to the tumor based on the comparison of the gene expression levels in the tumor to the reference levels comprises using an algorithm to calculate a score.

14. The method of claim 4, wherein the NF-kB activator is an anticancer agent, preferably selected from the group consisting of taxanes, vinca alkaloids, and topoisomerase inhibitors.

15. The method of claim 4, wherein the NF-kB inhibitor is selected from the group consisting of sulfasalazine, Luteolin, rapamycin, temsirolimus and everolimus, caffeic acid phenethylester, SN50, parthenolide, triptolide, wedelolactone, lactacystin, substituted resorcinols, (E)-3-(4-methylphenylsulfonyl)-2-propenenitrile, Bay 11-7082, Bay 11-7821, or Bay 11-7085, Pranlukast, etoposide, bortezomib, MLN9708, PS-1145, tetrahydrocurcuminoids, such as Tetrahydrocurcuminoid CG, extracts of Paulownia tomentosa wood, and MG-132 (Z-Leu-Leu-Leu-H).

16. The method of claim 4, wherein the immunotherapy is selected from the group consisting of administration of dendritic cells or peptides with adjuvant; DNA-based vaccines; cytokines; cyclophosphamide; anti-interleukin-2R immunotoxins; and antibodies, virus-based vaccines, formulations of Toll-like Receptor or RIG-I-like receptor ligands, Adoptive T cell therapy or other cell types.

17. The method of claim 16, wherein the antibodies are selected from the group consisting of anti-CD137, anti-PD1, anti-CD40, anti-PDL1, and anti-CTLA-4 antibodies.